100% found this document useful (1 vote)
291 views420 pages

Accounting and Finance For Managers

Contabilidad

Uploaded by

dario
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100% found this document useful (1 vote)
291 views420 pages

Accounting and Finance For Managers

Contabilidad

Uploaded by

dario
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 420

Note on the Ebook Edition

For an optimal reading experience, please view largetables and figures in landscape mode.

This ebook published in 2014 by

Kogan Page Limited


2nd Floor, 45 Gee Street
London EC1V 3RS
UK
www.koganpage.com

© Matt Bamber and Simon Parry, 2014

E-ISBN 978 0 7494 6914 6

Full imprint details


Contents

Introduction

01 Introduction to accounting
Objective
Learning outcomes
Key topics covered
Management issues
Introduction
Who and what is an accountant?
The two forms of accounting: financial accounting and management accounting
The financial accountant
Who are the users of financial accounting information?
The regulatory and conceptual framework
The qualitative characteristics of useful financial information
The regulatory framework
The standard-setting process
Types of business entity
The annual report and financial statements
The elements of financial statements
The articulation of the financial statements
A brief guide to some key problems and issues with financial statements
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

02 Accounting concepts and systems


Objective
Learning outcomes
Key topics covered
Management issues
Introduction
What is the purpose of the financial statements?
Statement of comprehensive income (income statement)
The statement of financial position (balance sheet)
The statement of cash flows
Preparing a set of financial statements
The income statement: cost of sales working
Underlying concepts: measurement rules and fundamental accounting concepts
Three further property, plant and equipment issues
Recording accounting information
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

03 Financial analysis: Part I


Objective
Learning outcomes
Key topics covered
Management issues
Introduction
Financial statement analysis for investment purposes
Other users and their needs
Horizontal analysis and trend analysis
Vertical analysis
Ratio analysis
Key ratios
Weaknesses and limitations
Conclusion
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

04 Financial analysis: Part II


Objective
Learning outcomes
Key topics covered
Management issues
Introduction
The drive for information
Stakeholder management
Corporate social responsibility reporting
Earnings announcements, conference calls and investor presentations
Media relations: press releases and newspaper coverage
Social media and internet bulletins
Conclusion
Comprehension questions
Answers to comprehension questions

05 Business planning
Objective
Learning outcomes
Key topics covered
Management issues
Introduction
Why budget?
Business planning and control: the role of budgets
The budget-setting process
Practical budget-setting
The basic steps of preparing a budget
Budgeting in different types of organization
Limitations and problems with budgeting
Improving business planning and budgeting
Conclusion
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

06 Budgets and performance management


Objective
Learning outcomes
Key topics covered
Management issues
Introduction
Responsibility centres
The controllability principle
Profit-related performance measurement
Standard costing
Standard costing and variance analysis
Variance analysis
Performance management in investment centres
Which is the best measure: ROI or EVA?
Non-financial performance indicators
The balanced scorecard
Performance measurement in not-for-profit organizations
Value for money (VFM) as a public-sector objective
NPO performance measurement: an example
Behavioural aspects of performance management: gaming and creative accounting
External influences on performance
Performance management in modern business systems
Conclusion
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

07 Cash flow
Objective
Learning outcomes
Key topics covered
Management issues
Introduction
Why does a business need cash?
What is cash flow?
How much cash does a business need?
Methods of establishing cash balances
Cash forecasting: the cash budget
Cash management: strategies for improving cash flow
Interpreting and analysing a cash-flow forecast
Conclusion
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

08 Pricing decisions
Objective
Learning outcomes
Key topics covered
Management issues
Introduction
The accountant’s perspective – costing and pricing
Absorption costing and full-cost-plus pricing
Problems with full-cost-plus pricing
Marginal-cost-plus pricing
Activity-based costing (ABC) pricing
Life-cycle costing and pricing
Conclusions: costing for pricing
The economist’s perspective
The marketer’s perspective
Combining the three perspectives: establishing an appropriate pricing strategy
Pricing strategies
Target pricing and target costing
Value engineering
Kaizen
Conclusion
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

09 Investment decisions
Objective
Learning outcomes
Key topics covered
Management issues
Introduction
Investment appraisal – the basics
Traditional evaluation techniques
Incorporating real-world complexities into investment appraisal
Investment appraisal within context
Taking a broader strategic view
Conclusion
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

10 Operational decisions
Objective
Learning outcomes
Key topics covered
Management issues
Introduction
Operational decision making
Cost–volume–profit analysis (CVP)
Relevant costing
Conclusion
Comprehension questions
Exercises
Answers to comprehension questions
Answers to exercises

Appendix A An introduction to double-entry bookkeeping


Appendix B International Accounting/Financial Reporting Standards (as at
02.04.2013)
Appendix C Example earnings announcements
Appendix D Discount tables
Appendix E Annuity factors
References

Index
Introduction

Ininformation
this book we hope to show you that, despite some imperfections and limitations, accounting
can help you make better-informed decisions as individuals and as managers.
Though we talk about accountants and accounting, we hope that you will begin to understand
and appreciate that this group of people do not produce all the accounting information. More
importantly, the information they produce needs to be interpreted at an operational, tactical and
strategic level by employees from across the organization regardless of background. You might
think that certain professions are naturally exempted from the push and pull of accounting, but
this is not true. Just ask any marketing director, for example, how he or she feels about this
year’s budget, or sales team managers what they think of their ‘on-target earnings’ bonus!
Despite the broad reach of the subject matter, accounting and finance as a discipline is often
viewed with unease as students commonly have a pre-existing belief that it requires a strong
grounding in mathematics, that it comprises a set of impossibly convoluted and complicated
rules, or that it is simply boring! But these presuppositions are unfounded and ultimately untrue.
What is true is that much of accounting and finance is a ‘building-block’ exercise ie one issue
sets the foundation upon which the next can be built. Therefore, we have provided exercises
throughout the text. Some of these are follow-my-lead ‘worked-through exercises’, while others
give you the opportunity to practise for yourselves. We urge you to attempt these. They provide
an opportunity to reflect on what you’ve learnt and question the underlying purpose of the
process and/or result. We have also given a number of real-life examples and illustrations to help
you get a better grasp of the material. We try to take you beyond the mechanics of generating ‘an
answer’ and towards an understanding of why that ‘answer’ might be important.
To make the book as accessible as possible we have divided it into two broad sections:
financial accounting (Chapters 1 to 4); and management accounting (Chapters 5 to 10). These are
the two major sub-disciplines of accounting and finance in professional and academic life. The
chapters vary quite considerably in length and to a certain degree this can be interpreted as a
reflection of our view of their relative importance to a subject novice. Throughout each topic, the
following advice pervades: the more you put in, the more you will get out. Only the tip of the
iceberg is revealed during each chapter and therefore we urge you to pursue other sources as far
and as wide as you can. That might be in the form of media releases, academic and trade
journals, discussion forums, and so forth. The deeper you engage, the more you will understand
and the greater the likelihood that you will ultimately be able to make better-informed decisions.
You might already be familiar with some of the more common terms such as ‘annual report’,
‘financial statements’, ‘accountant’ and ‘auditor’, but you might not know what they mean. That
is fine. The financial accounting section expands on these and other related topics. The
management accounting section introduces some practical and useful tools, methods and
techniques to help you deliver solutions and suggestions to some major business problems.
The financial accounting section presents some common business vehicles, ie sole traders,
partnerships, limited liability companies and public limited companies. The advantages and
disadvantages of each are subsequently explained and explored. Though the first few chapters
focus on corporate reporting practices (as it is simpler to extrapolate these to simpler forms of
accounting than the other way round), much of the information and advice provided within both
the financial accounting and management accounting chapters can also be mapped to other
organizations, including those within the not-for-profit sector.

How to approach the content


Our objective has been to include key (or core) information in the body of the text. Alongside
this, we have also included a number of examples and exercises. If you follow the primary
exercise on the first run through, that is excellent and you might like to simply move on. Many,
however, won’t and if you fall into this camp there is no harm in following it through in more
detail in your own time. We have also provided some supporting examples which run alongside
the main text. These secondary and supplementary examples are different enough to stop even
the most advanced students from becoming bored. It has always seemed strange to us that
accounting and finance are portrayed as subjects which are passively learnt, on a surface level,
facilitated by a teacher-led approach. We do not concur. You should see these examples the same
way as you might learn to catch a ball, take a penalty, undertake an experiment, choose the right
chemical etc. The examples help you to grow in experience and guide your future decision-
making behaviour. Indeed, we have an advantage over other academic disciplines as our
examples are interesting because they are real! You can see them in your everyday life, often
panning out before your eyes. Once you begin to understand the subject, you will never see the
world in the same way again.
We have also included notes which we have called ‘expert view’. The idea behind these is to
encourage you to think more broadly and present you with further issues and questions that you
could consider pursuing. Though these are non-essential, they bridge the gap between this
introductory text and other, more advanced syllabuses. Solutions in accounting are not binary as
many uninitiated believe. There is plenty of scope for discretion in practice and behaviour
(normally legal but occasionally illegal!). Even at this first step into the discipline, we hope to
show you some interesting arguments, conflicting viewpoints and how they (have) arise(n).

A note on terminology
We feel it is worth making clear at this early stage that this is a textbook designed for an
international audience. Throughout, therefore, we have adopted International Financial Reporting
Standards (IFRS) terminology. Occasionally we have referenced the relevant document produced
by the International Accounting Standards Board (IASB), eg an International Financial
Reporting Standard (IFRS), an International Accounting Standard (IAS) or the Conceptual
Framework (F). Though we don’t recommend beginners to pursue these references, there are
some who might want to. If you are interested, a site which is well worth a visit and which we
strongly recommend is IASplus.com. This site is maintained by one of the Big Four1 accounting
firms – Deloitte. It is generally recognized as the go-to site for international financial reporting
updates.

Proposed course outline


We fully expect lecturers who adopt this text to adopt different approaches to the content and the
method of delivery. The book is designed to aid a standard ‘lecture series followed by tutorial’
approach, but the material and exercises have also been trialled in workshops and smaller
seminar groups and found suitable for all sizes and formats.
University course lengths differ by institution and jurisdiction. Some standard course outlines
are provided below.

Twenty-hour course (10 weeks * 2-hour lectures)


Week 1 Chapter 1
Week 2 Chapter 2
Week 3 Chapter 3
Week 4 Chapter 4
Week 5 Chapter 5
Week 6 Chapter 6
Week 7 Chapter 7
Week 8 Chapter 8
Week 9 Chapter 9
Week 10 Chapter 10

Sixteen-hour course (8 weeks * 2-hour lectures)


Week 1 Chapter 1
Week 2 Chapter 2
Week 3 Chapter 3
Week 4 Chapter 4
Week 5 Chapters 5 & 10
Week 6 Chapters 6 & 7
Week 7 Chapter 8
Week 8 Chapter 9

Twenty-four-hour course (24 weeks * 1 hour lecture; or 12


* 2-hour lectures)
Week 1 Chapter 1
Week 2 Chapter 1
Week 3 Chapter 2
Week 4 Chapter 2
Week 5 Subsidiary bookkeeping appendix
Week 6 Subsidiary bookkeeping appendix
Week 7 Chapter 3
Week 8 Chapter 3
Week 9 Chapter 4
Week 10 Chapter 4
Week 11 Chapter 5
Week 12 Chapter 5
Week 13 Chapter 6
Week 14 Chapter 6
Week 15 Chapter 7
Week 16 Chapter 7
Week 17 Chapter 7
Week 18 Chapter 8
Week 19 Chapter 8
Week 20 Chapter 9
Week 21 Chapter 9
Week 22 Chapter 9
Week 23 Chapter 10
Week 24 Chapter 10

Note
1 Deloitte, PricewaterhouseCoopers, Ernst & Young (EY), KPMG.
01
Introduction to accounting

OBJECTIVE
To provide an overview of the conceptual and regulatory framework that underpins financial accounting and an
understanding of the content and structure of the financial statements.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Discuss the role of the financial accountant and the information which they prepare.
Describe the relationship between the statement of financial position, income statement and statement of cash flows.
Understand the elements and structure of the annual report and primary financial statements.
Describe the conceptual and regulatory framework of accounting.
Evaluate what information the financial statements cannot provide.

KEY TOPICS COVERED


Why do businesses need financial accountants and financial statements?
The type of information financial accounts record.
The concepts of assets and liabilities, income and expenditure.
Introduction to the three primary financial statements: the statement of financial position, income statement and the
statement of cash flows.
Introduction to the accounting regulatory system and the way it has shaped the informational content of financial
statements.
Discussion of some key problems and issues with financial statements.

MANAGEMENT ISSUES
Managers need the skills to be able to ‘read’ rather than prepare financial statements. Equally importantly, they need an
understanding of what financial statements can and cannot tell them about a business.
Introduction
Though it is incredible to us, there are always a minority of newcomers to accounting and
finance who believe that our discipline is dull! Yet nothing could be further from the truth. Our
job is perennially fascinating because it is dynamic. The role is constantly being redefined by
innovation. The information you are about to learn has the ability to guide individual and
corporate decision making fundamentally at every level. If you (choose to) work in an
accounting firm or in industry, then rest assured that these places are almost always vibrant,
energetic and intellectually stimulating.
Chapters 1–4 of this book discuss the role of the financial accountant alongside the purpose
and usefulness of the information they produce. We outline some basic preparation rules,
propose several analysis and interpretation methods, and introduce some broader questions about
financial accounting and its role. Chapters 5–10 are concerned with management accounting and
the sub-discipline financial management.
In this chapter we open by addressing some of the questions most frequently asked by those
who are new to the world of accounting and finance, namely: What is an accountant? What is the
accountant’s role? What is the broad purpose of this function in an organization? We introduce
the key elements of the conceptual and regulatory frameworks and follow this by identifying the
most common types of business vehicle. This chapter therefore outlines who the preparers and
users of financial information are. In addition, we introduce the standard-setters and describe
their role in the accounting community. We conclude this chapter by providing an overview of
some of the commonly perceived key problems associated with the main information product
output of the financial accountant: the annual report and financial statements.
Having taught this for many years, we are well aware that there are learners who presume this
material is both turgid and irrelevant. This combination is fatal to engagement. We have
therefore tried to make this as interesting as possible by breaking up the delivery of ‘hard facts’
by bringing in other issues, such as who uses the information and why it might be important
(necessary?) for there to be a legitimate and credible standard-setting body. A further problem is
that this material can feel disjointed. This is because there is a minimum amount that you need to
know and, sadly, there isn’t enough space in an introductory textbook to consider some of the
more interesting issues that underlie its development, refinement, purpose and importance. In an
introductory text it seems excessively cruel to put forward these questions where there are no
right or wrong answers. We would, however, like to take this opportunity to ask you to consider
them as you make your way through the material. For example: Do you think a global financial
reporting system is desirable? Do you think it is right that accountants should be allowed to be
self-governing? Do you think domestic governments should intervene to make rules fairer to
their jurisdiction? Do you think accounting systems should comprise a set of restrictive rules or
do you think they should be based on principles? Why is there a vibrant accounting and finance
academic community? Additionally, given the rapid pace of change in practice (in real life), how
can the relatively slow-moving research community help the profession?
Who and what is an accountant?
One thing that has remained unchanged for millennia is that every day, in our private and work
lives, we rely on accountants and accounting information to make better-informed decisions. To
put this in context, some examples of decisions you might face include whether to: buy an asset,
rent it, or make it; price a product higher or lower than your competitors; outsource a product,
project, department or operation; take on a new contract; acquire a competing business. Needless
to say, there are many more!
The Oxford dictionary defines accounting (noun) as ‘the process or work of keeping financial
accounts’ while an accountant (noun) is ‘a person whose job is to keep or inspect financial
accounts’. This definition, however, is overly simplistic. For a long time, accounting was
considered a process of collecting, analysing and communicating financial information to allow
users to make better-informed decisions. This work remains at the forefront of the role but more
recently accountants have been asked to expand their remit into new areas. For example,
accountants are now deeply involved in the preparation of non-financial information, including
corporate social responsibility reporting. The image of the staid, conservative, grey-suited drone
is outdated. The job has changed. We now need to be communicators as well as doers. We need
to be client facing rather than just hit buttons on a calculator. There is a new breed of accountant
and whether you believe this is a change for the better or worse, it seems to be a change that is
here to stay.
The good news is that accounting is largely a refinement of common sense. Among other
factors, relative levels of reliance on financial information, increased regulation and
professionalization have driven increased sophistication of presentation and methods, but at the
heart of all accounting is the idea that financial data should be converted into useful information
for decision making. If either the process or the output were to become overly complex or
burdensome, the profession might risk damaging this core objective.

Expert view 1.1: Seeking definitions – what is an accountant?


The subtleties and complexities of the definition problem are illustrated by the recently drafted consultation
paper produced by the International Federation of Accountants (IFAC) in 2011, entitled ‘A proposed definition
of “professional accountant”’. This work drew attention to a number of important problems exacerbated by a
lack of specificity in the term ‘professional accountant’.
IFAC organized their definition into three descriptive levels:

1. Initially, the term ‘professional accountant’ is defined by emphasizing some form of official accounting
qualification (eg formal education, certification, or chartering).
2. IFAC then state what a professional accountant does by outlining the core responsibilities that imply the
application of skills in the context of society’s expectations. Thus, their definition recognizes the public-
interest responsibilities of accountants and their profession.
3. Finally (which is optional and contingent upon the characteristics of each jurisdiction), the definition states
that professional accountants can be differentiated from one another by certain factors, such as types of
responsibilities (eg public-sector accounting, auditing role) and the level of formal training or education
generally.
A simple way to understand the role of an accountant is through an illustration. As you work
through the paragraph, try to picture yourself as the protagonist. You will see how and why the
simplest forms of accounting developed.

ILLUSTRATION 1.1 Climb On!


For a long time, you have been wondering how to convert your passion for rock climbing into a business opportunity.
During a recent climbing trip you met Chris, a climbing-gear designer. He agreed to give you 100 chalk bags and 100
climb-oriented t-shirts for $4 and $5 each, respectively. You came to an agreement with him that you’d make
payments on an ad-hoc basis as the goods were sold. As soon as you arrived home you listed the first 30 chalk bags
and 50 t-shirts on an internet-based auction site. These sold within three days of their listing for $8 and $10
respectively. Half of these customers paid immediately. You offered 20-day credit terms on the sales and your past
experiences tell you that customers tend to take full advantage of this policy.
You are keen to work out your financial position before you proceed any further.

Table 1.1 Solution to Illustration 1.1

What do you own? What do you owe?


You owe Chris money for the 100 chalk
70 chalk bags which you bought for $4 each $280 $400
bags at $4 each
You owe Chris money for the 100 t-shirts
50 t-shirts which you bought for $5 each $250 $500
at $5 each
You received $8 per chalk bag, and collected 15 units worth of
$120
sales immediately
You are owed for (the remaining) 15 chalk bags sold at $8 per
$120
unit
You received $10 per t-shirt, and collected 25 t-shirts worth of
$250
sales immediately
You are owed for (the remaining) 25 t-shirts sold at $10 per
$250
unit

Through the medium of this example, we have introduced many of the basic tenets of
accounting. Your mind should be racing with the same questions as business owners and their
stakeholders have been asking for millennia. From an internal perspective, you were simply
interested in your financial position:

How much more (or less) wealthy am I?


Who do I owe money to?
How much do I owe?
Who owes me?
How much do they owe me?
How much cash have I got?
How much are my inventories (unsold goods) worth?

However, given this analysis your interest might have been piqued and you are now pondering
other associated questions about your performance, future position and cash flows, plus any
constraints and opportunities which you might face. These might include:

Have I excluded any costs?


Have I made a profit?
Am I charging a sensible price?
Will all my customers pay?
How long will Chris be willing to wait for his payment?
Will Chris allow me to buy more inventories?
Can I negotiate the purchase price down?
Will the goods Chris sells me be manufactured to the same quality?
… and so forth.

Extract 1.1: Perspectives from accounting research


Many people who first start studying accounting and finance cannot believe that there is a large, thriving, active
research community. The problem is that more often than not, all that people know of accountants comes
through the media who like to portray us, and the information we produce, in binary terms. They lead casual
audiences to believe that it is ‘right’ or ‘wrong’, ‘black’ or ‘white’. What they don’t tell you is that the discipline is
a sea of grey (not grey suits) and rarely are decisions so clear cut.
Maintaining and cultivating the link between practice (ie the accounting profession) and academia is
important. There needs to be a vibrant academic community in the same way that the academic community
needs a strong, credible and socially responsible accounting profession. There are those who argue that the
gap has never been wider, while others argue that it’s never been narrower. Kaplan (2011) has a more
balanced view and suggests in his paper, ‘Accounting scholarship that advances professional knowledge and
practice’, that though research (academic scholarship) has helped craft the direction of professional decision
making and contributed to professional knowledge, there are many areas where the community could do more.
He states:

As accounting scholars have focused on understanding how markets and users process accounting data,
they have distanced themselves from the accounting process itself. Accounting scholarship has failed to
address important measurement and valuation issues that have arisen in the past 40 years of practice. This
gap is illustrated with missed opportunities in risk measurement and management and the estimation of the
fair value of complex financial securities.
Kaplan (2011)

Let’s not be too disheartened, though, for when a community recognizes a performance gap and there is a
positive attitude towards change, this is normally the first tentative step on the pathway to finding solutions.

The two forms of accounting: financial accounting and


management accounting
Broadly speaking, accounting divides into two forms: financial accounting and management
accounting. The principal differences can best be summarized into four categories as shown in
Table 1.2.

Table 1.2 Differences between financial accounting and management accounting

Key
Financial accounting Management accounting
differences
External, principally owners of
Users Internal management of the company
the business eg shareholders
Governed by region-specific law,
Format exchange-specific regulation and Can take any form depending upon the purpose of the information
accounting standards
Normally annually. As required.
Large listed entities sometimes Some information is produced and some exercises are performed, as
Frequency
report quarterly or half-yearly a matter of course, daily, weekly, monthly, quarterly and annually.
dependent upon regulation Other information is required on an ad-hoc basis
Dominated by historic information Forward-looking perspective. Detailed analysis of past results and
Content
based on past transactions incorporation of known changes, in order to plan for the future

Exercises: now attempt Exercise 1.1 on page 39

Expert view 1.2: The origins of accounting


Bookkeeping, as we know it, was first documented by Luca Pacioli (often referred to as the ‘Father of
Accounting’) in his text Summa de arithmetica, geometria, proportioni et proportionalita (Venice, 1497). Within
this mathematics textbook was a section describing the method of double-entry bookkeeping employed by
Venetian merchants. Pacioli described a system of journals and ledgers which accounted for assets, liabilities,
gains and losses. However, systems of recording assets and liabilities, income and expenditure have existed
for millennia. For example, clay tablets which appear to record financial transactions have been found in Egypt
and Mesopotamia dating to before 2000 BC. Indeed, Jones (2011) dates the first accounting scandal to the
same period.

The financial accountant


So, we know there are fundamental differences between financial accounting and management
accounting. Let’s now look more closely at where one would find financial accountants and what
they are responsible for producing and preparing. It is common for financial accountants to work
within accounting practices that provide external services and advice to entities about their
financial reporting and related matters, eg audit and accounting services. They can also be found
within organizations, preparing in-house accounting records and associated information.
Expert view 1.3: Professional accounting qualifications
To the uninitiated, it might be natural to presume that when people refer to themselves as ‘an accountant’, they
mean ‘Chartered Accountant’. However, to be able to call oneself a Chartered Accountant it is necessary to be
a member of the Institute of Chartered Accountants in England and Wales (ICAEW), Scotland (ICAS), Ireland
(CAI), Australia (ICAA), Zimbabwe (ICAZ), New Zealand (ICANZ), The Canadian Institute of Chartered
Accountants (CICA) or The South African Institute of Chartered Accountants (SAICA).
Job adverts in the UK commonly specify ‘CCAB qualified accountant required’. The Consultative Committee
of Accountancy Bodies (CCAB) is an umbrella group and has five member bodies: Association of Chartered
Certified Accountants (ACCA); Chartered Institute of Public Finance and Accountancy (CIPFA); ICAEW; CAI;
and ICAS.

NOTE: The Chartered Institute of Management Accountants (CIMA) gave notice to withdraw from this group in
March 2011.

It is common for financial accountants to be trained by, and work within, accounting
partnerships. These range from sole practitioners working in isolation to global organizations
generating tens of billions of dollars in revenue per annum. There are a number of so-called elite
professional accounting firms which have become household names. Collectively they are
known as the Big Four, essentially because there are four of them: PricewaterhouseCoopers;
Deloitte; KPMG; and Ernst and Young (now known as EY).
There are fears about the long-term viability of this dominance and most in the profession
expect changes in the not too distant future (see Extract 1.2). However, the accounting services
marketplace is competitive. There are hundreds of accounting firms outside of these Big Four.
Organizations regularly put work out for tender as they are not obliged to stay with one firm.
Indeed, from a commercial perspective, there is no better way of being offered a lower fee from
your current auditor than by asking others to bid competitively against them!

Extract 1.2: Big Four dominance

UK Competition Commission Says Big Four Audit Dominance Not


Best for Investors
March 4, 2013
Accountingweb
By Frank Byrt

The United Kingdom’s Competition Commission (UKCC) says that the nation’s audit market is dominated by the Big
Four accounting firms, which has stanched competition for audit work from public companies to the detriment of their
shareholders. The UKCC, in its ‘Statutory Audit Services Market Investigation’ report released on 22 February 2013,
suggested that among the possible remedies is mandatory rotation of audit firms.
UKCC’s report is a major milestone in a 16-month probe that was set in motion by a critical report from the
House of Lords Economic Affairs Committee in 2011. The UKCC’s inquiry focused on Big Four firms KPMG,
Deloitte, Ernst & Young (EY) and PricewaterhouseCoopers (PwC).
‘Essentially, we identified two clusters of issues,’ UKCC Audit Investigation Group Chair Laura Carstensen
told AccountingWEB UK. ‘The first was “stickiness” and propensity of companies not to switch auditors and
adverse issues that can result. And the second was to make sure auditors are more squarely aligned with what
shareholders want.’

UKCC Recommendations

Mandatory tendering.
Mandatory rotation of audit firms.
Expanded remit and/or more frequent audit quality reviews.
Prohibit ‘Big Four only’ clauses in loan documentation.
Strengthen accountability to the audit committee.
Better shareholder–auditor engagement.
Extended reporting requirements.

The UKCC concluded that Big Four firms hold most of the big company audits and that those organizations rarely change
auditors, which hurts the competition for public company audit work and results in higher prices, lower quality, and less
innovation and differentiation than would be the case in a more open market.
The lack of competition creates a risk of auditors being insufficiently independent from executives and
insufficiently sceptical of their attempts to present the accounts in the best possible light, the report said.

How can this be interpreted and why is it relevant to you?


The existence of a Big Four has benefits but there are also drawbacks. Extract 1.2 identifies two potential issues:
stickiness; and a threat to goal congruence. There is some debate about whether audit firms should be forced to rotate
after a fixed term has been served, ie the job should be mandatorily put out to tender every five years, maybe with the
incumbent firm not being allowed to engage in the process. The former has recently been proposed in the UK by the
Competition Commission in a series of measures to limit the dominance of these élite firms and increase competition in
the market. It is clear, however, that though this will solve one set of problems, it creates another: eg set-up costs, learning
curve effects, a risk that sector/industry audit specialism won’t be developed. There is no doubt that an audit firm who ‘is
in the pocket’ of the senior executive (or vice versa) is an unhealthy situation, but none of the proposed solutions is
perfect.

Ultimately, the central function of the financial accountant is to prepare (review or audit) the
annual report and financial statements. Traditionally this document acts as a summary of the
organization’s performance and position. More recently it has become a repository for other
information, and this point will be discussed in more detail later.
Alongside providing accounting support, the range of services provided by financial
accountants (an accounting firm) can be summarized as follows:

assurance services, including auditing and regulatory compliance;


taxation services, including taxation planning, developing taxation strategies and taxation
compliance work;
transaction work, including advice to help businesses grow, prosper or reinvent themselves
on issues such as obtaining funding, or mergers and acquisitions;
advisory services, including internal audit, risk management advice and corporate recovery
(insolvency).

Who are the users of financial accounting information?


If financial accountants are responsible for the preparation of the annual report and financial
statements, it is worth asking the follow-on questions: ‘who uses this information?’ and ‘what do
they use this information for?’ (Figure 1.1).

Figure 1.1 Users of financial reporting information

The Conceptual Framework for Financial Reporting (Conceptual Framework or ‘F’) identifies
the primary users of general-purpose financial reporting as present and potential investors,
lenders and other creditors. They use this information to make decisions about buying, selling or
holding equity or debt instruments and providing or settling loans or other forms of credit [F,
objective {OB} 2]. The information is intended to allow these stakeholders to assess an entity’s
performance and position as well as gauge future cash-flow prospects. These users will probably
be keen to assess how well management have discharged their responsibilities in the use of the
entity’s resources [F, OB4].
The Conceptual Framework notes that there are also other stakeholders who might find
general-purpose financial reports useful. Among these are regulators, the public, the government,
customers, employees and competitors. The IASB notes that interested parties should not limit
themselves to the financial report; there is other information available which would be useful
when undertaking a full assessment of an entity’s position and performance.

Exercises: now attempt Exercise 1.2 on page 39

The regulatory and conceptual framework


There are a limited number of professions that can claim to be self-governing, but accounting is
one of them. The government has ceded responsibility to a group of experts whom they have
deemed to be credible, reliable and socially responsible. In many ways, this is a great benefit. It
does mean, however, that the central coordinating body (or bodies) need to be able to respond to
any problems quickly and efficiently and with the requisite levels of ability and skill when
challenged by economic, political, social and technological complexities.
The profession needs a strong, robust and fair conceptual and regulatory framework. The
following paragraphs aim to introduce the main players in the environment as well as outline
some of the processes.

Expert view 1.4: A current problem for the accounting standard-setters


A recent example of the type of problem faced by the profession is the case of financial instruments reporting
standards. These are being regularly updated to take account of financial developments in the use and
complexity of derivative financial instruments, eg what value should these be shown at: cost or market value?
This decision is made more difficult when you know that the cost of these commonly economically significant
assets/liabilities is normally $nil (or close to $nil). Standard-setters need to ask themselves: ‘How do you value
an instrument where there is no market?’ ‘And how do you create a rule that would be fit for all purposes?’
‘How do you allow management to explain to the reader the risks faced by the entity as a result of holding
these instruments?’ ‘Is there a minimum amount they should disclose?’ It is amazing to think that a decade or
so ago, companies were able to hold massive amounts of highly volatile derivative financial instruments without
it being mandatory to declare how much they were worth. Though the rules we have now may not be perfect,
they are certainly more useful than what we had before.

Principles-based versus rules-based standards


The IASB, in their role as standard-setter, aspire to create principles-driven requirements rather
than rules-driven regulations. This is thought to alleviate some of the problems created by hard
rules (eg de jure compliance, or ‘work around the rules’) and allow organizations to be able to
meet requirements even when faced with a situation not previously encountered. This does,
however, mean that organizations frequently need to make judgements because the reporting
code is not completely rigid and immovable.

The Conceptual Framework


In addition to the way the regulations are written, it is also important to note that there is a
document outlining the Conceptual Framework for accounting. This provides supplementary
guidance when there are no specific reporting requirements or regulations governing the
reporting area. That means that this document can be used to guide preparers when there isn’t a
standard that tells them what to do. The Conceptual Framework is a frame of reference which has
two purposes: first, it is used by companies to aid their financial reporting decision making; and
second, it is used by the IASB to help them to develop new accounting standards and revise old
ones.
The Conceptual Framework deals with the following key issues:

a the objective of financial reporting;


b the qualitative characteristics of useful financial information;
c the definition, recognition and measurement of the elements from which financial statements
are constructed; and
d concepts of capital and capital maintenance.

The objective of general-purpose financial reporting is to provide financial information about the
reporting entity that is useful to key stakeholders. The concept of usefulness is subject to great
debate. Accounting and finance research has attempted to define and refine the notion of
financial reporting quality. The key problem, however, is that quality is a nebulous term and thus
the integrity and robustness of any definition is questionable.

The qualitative characteristics of useful financial


information
Defining a nebulous concept such as ‘quality’ has proved to be difficult for everyone who has
ever approached the subject. For example, how easy is it to say what makes Shakespeare’s
Hamlet such a wonderful play? What contributes to its quality? Can those characteristics be
mapped to other texts? Accountants and their representative bodies have had the same problem.
What characteristics ensure that information is of sufficient quality; or useful for decision
making? In response, the IASB state that for financial information to be useful, first and
foremost it must be relevant and faithfully represent what it purports to represent. There are then
four further enhancing characteristics: comparability, verifiability, timeliness and
understandability. These are summarized in Figure 1.2 and some definitions of these
characteristics follow.

Figure 1.2 The qualitative characteristics


It is almost impossible to define quality, but it is also reasonably difficult to define these other
terms. We thought it would be useful to set out how the standard-setters have chosen to define
them (F, Qualitative Characteristics {QC} 6–20):

The fundamental characteristics


Relevant financial information is capable of making a difference in the decisions made by
users. Information may be capable of making a difference in a decision even if some users
choose not to take advantage of it or are already aware of it from other sources. Financial
information is capable of making a difference in decisions if it has predictive value,
confirmatory value or both.
Financial reports represent economic phenomena in words and numbers. To be useful,
financial information must not only represent relevant phenomena, but it must also
faithfully represent the phenomena that it purports to represent. To be a perfectly faithful
representation, a depiction would have three characteristics. It would be complete, neutral
and free from error.

The enhancing characteristics


Users’ decisions involve choosing between alternatives, for example, selling or holding an
investment, or investing in one reporting entity or another. Consequently, information about
a reporting entity is more useful if it can be compared with similar information about other
entities and with similar information about the same entity for another period or another
date. Comparability is the qualitative characteristic that enables users to identify and
understand similarities in, and differences among, items.
Verifiability helps assure users that information faithfully represents the economic
phenomena it purports to represent. Verifiability means that different knowledgeable and
independent observers could reach consensus, although not necessarily complete agreement,
that a particular depiction is a faithful representation. Quantified information need not be a
single-point estimate to be verifiable. A range of possible amounts and the related
probabilities can also be verified.
Timeliness means having information available to decision-makers in time to be capable
of influencing their decisions. Generally, the older the information is the less useful it is.
However, some information may continue to be timely long after the end of a reporting
period because, for example, some users may need to identify and assess trends.
Financial reports are prepared for users who have a reasonable knowledge of business and
economic activities and who review and analyse the information diligently. Classifying,
characterizing and presenting information clearly and concisely make it understandable.

Materiality
You will often hear accountants refer to the separate notion of materiality. This concept is
critical to an understanding of the aims and objectives of financial reporting and auditing. An
item is deemed material if its omission or misstatement could influence decisions that users make
on the basis of financial information about a specific reporting entity.
Materiality therefore is an entity-specific variable. It is based on the nature or magnitude (or
both) of the item(s) to which the information relates in the context of an individual entity’s
position. Consequently, there is no single specified measure or quantitative threshold for
materiality. Each item must be reviewed on its own merits and in the broader context.

Exercises: now attempt Exercise 1.3 on page 40


The regulatory framework
We have briefly introduced the Conceptual Framework and therefore it is now time to turn to the
regulatory framework. It is not essential at this level of study for you to know the intricacies of
the accounting standard-setting process or the wider regulatory framework. However, an outline
of the structure of the major bodies and processes form both interesting and worthwhile
knowledge. An outline of the structure of the IFRS Foundation is shown in Figure 1.3. Figure 1.4
provides a brief overview of the phases of the standard-setting due process.

Figure 1.3 The structure of the IFRS Foundation (www.IFRS.org)

Figure 1.4 IASB standard-setting due process


The IFRS Foundation
The IFRS Foundation is an independent, not-for-profit private-sector organization. The IFRS
Foundation sets at its core the notion of public interest which is mirrored by the definition of
accountants and their work (eg IFAC, 2011). The principal objectives of the IFRS Foundation
are as follows:

to develop a single set of high-quality, understandable, enforceable and globally accepted


International Financial Reporting Standards (IFRSs) through its standard-setting body, the
IASB;
to promote the use and rigorous application of those standards;
to take account of the financial reporting needs of emerging economies and small and
medium-sized entities (SMEs); and
to promote and facilitate adoption of IFRSs, being the standards and interpretations issued
by the IASB, through the convergence of national accounting standards and IFRSs.

Source: www.IFRS.org

The IASB
The IASB acts as an independent standard-setting body on behalf of the IFRS Foundation. The
IASB currently has 15 full-time members who are responsible for the development and
publication of IFRSs. Interestingly, all meetings of the IASB are held in public and are webcast.
The IASB and the IFRS Foundation heavily stress the notions of comparability, inclusiveness
and consultation. They actively engage with stakeholders from around the world, including
investors, analysts, regulators, business leaders, other accounting standard-setters and the
accountancy profession.

The IFRS Interpretations Committee


The 14 members of the IFRS Interpretations Committee are responsible for reviewing, on a
timely basis, widespread accounting issues that have arisen within the context of current IFRSs
and to provide authoritative guidance (IFRICs) on those issues. The members are drawn from
different professions and countries.

Expert view 1.5: Is this an unfair or biased system?


Though there is research which finds limited evidence of varying levels of relative influence among stakeholder
groups during the standard-setting process, the IASB have put in place a system which has the potential to be
fair and rigorous (see Figure 1.4). There is a combination of observable and non-observable phases and
though it has been argued that these ‘dark’ periods impair transparency, they also allow the Board to take,
behind closed doors, valuable expert advice which otherwise might not be provided. In order to make the
process more transparent, the Board has recently opened up Board meetings to the public, made minutes
available and also webcasts them.

Exercises: now attempt Exercise 1.4 on page 40

The standard-setting process


Figure 1.4 outlines the standard-setting process in financial reporting. This is the template for
accounting requirements’ innovation and implementation. As you can see, the process is
designed to be robust, involved and transparent as the outcome needs to be agreed among
stakeholders. Agreement between constituents – including preparers, users, auditors,
representative bodies and so forth – helps to ensure the continued legitimacy of the standard-
setter. It should also mean that the requirements are likely to be met by preparers and their value
understood by users.

Types of business entity


We are now at the crossroads between topics being covered in this chapter. The first part
considered the role of the accountant and the systems and processes that have grown up around
them. The conceptual and regulatory framework has been described. For accounting to exist,
there must be something to account for. Therefore in terms of basic background information, we
address two further questions: ‘What is a business?’ and ‘What is the purpose of a business?’
The answer to both questions is far from straightforward. Indeed, they often overlap and
intertwine.

What is a business?
One could argue that in its simplest sense a business should be defined by its legal form.
Therefore, we shall briefly review the most common types of business entity that exist.
There are a wide range of business vehicles, or forms of business ownership, available. Each
has their relative advantages and disadvantages. In the UK, the government has developed an
excellent resource which explains the major differences between the various business vehicles
available. The site also offers advice on how one might set up a business
(https://www.gov.uk/browse/business/setting-up). There are three which are particularly useful to be
aware of:

1. sole trader;
2. partnership;
3. limited company.

Sole trader
As the name suggests, a sole trader is someone opting to work for themselves. As seen in the
example above – Climb On! – we put ‘you’ in the position of running your own business. The
business is not incorporated, ie it is not a limited liability company. You do not share the
ownership of the business with anyone else, ie you have no business partner(s). This doesn’t
mean that you cannot have any employees, it is just not common. Businesses of this type are
normally quite small in terms of assets, liabilities, revenues and expenses. There are many
everyday examples of people going into business on their own, such as plumbers, electricians,
hairdressers, private tutors, artists, photographers and so forth.
The financial accounting information a sole trader is required to produce is limited. It is
normally required to satisfy tax authorities’ purposes. Occasionally financial institutions might
request specific information, particularly for lending purposes.
Establishing oneself as a sole trader, therefore, is frequently the first step for many
entrepreneurs. The main advantage of this business vehicle is the retention of control. The owner
is solely responsible for all decision making. The other key advantage is the reduced accounting
and legal regulatory burden. The major disadvantage is the unlimited liability! In other words, a
sole trader is wholly responsible for all liabilities of the business.

Partnership
A partnership occurs when two or more people decide to run a business together. As with
establishing a business as a sole trader, the regulatory burden is low, especially in comparison
with limited companies. Partnerships range considerably in size. It is probably more common
that they have low assets, liabilities, revenues and expenses; however, they can be huge global
businesses. Examples again include hairdressers, plumbers, tutors and so forth. They also include
businesses such as medical practices, legal practices and accounting practices.
The advantage of forming a partnership is that you can share responsibility and the burdens of
ownership. It is also likely that the skill-set available will be more varied and expertise could be
easier to channel towards specific projects. In an accounting partnership, for example, you might
want the skills of both a financial accountant and a tax accountant; it is rare that one person is an
expert in both. The disadvantages largely stem from behavioural issues. Sharing ownership often
places significant burdens on pre-existing well-functioning relationships. Also, there are some
people who simply like making all the decisions, not sharing this role!

Limited company
In the UK, a privately held business is referred to as a limited company; this is commonly
abbreviated to ‘Ltd’. The equivalent term used for a private company in Australia is Proprietary
Limited Company (abbreviated to Pty. Ltd.). In India and Pakistan the designation Pvt. Ltd.
(Private Limited) is used for all private limited companies. In South Africa the term Pty. Ltd. is
used. In the United States, the expression ‘corporation’ is preferred to limited company and it is
common to see the abbreviation ‘Inc.’ (short for ‘Incorporated’), but in many states ‘Ltd.’ is also
permitted.
The word ‘limited’ relates to the level of financial exposure. An entity can be incorporated as
a limited liability company, at which point, in law, it becomes a separate legal entity. Therefore,
while sole traders and partners are personally responsible for the amounts owed by their
businesses, the shareholders of a limited liability company are only responsible for the amount to
be paid for their shares. A limited liability company conducts all activities in the name of the
entity, eg invoices are issued in the company’s name, bank accounts need to be set up in the
company’s name (not directors’ or managers’ names). Note, however, that it is not uncommon
for lenders and trading partners to ask the directors of companies for personal guarantees, which,
of course, negates much of the advantage associated with incorporation.
As with partnerships, companies can range in size from small to huge. Most companies which
are household names are public limited companies. There are several advantages and
disadvantages to incorporating. Some of these are shown in Table 1.3.

Table 1.3 Advantages and disadvantages of incorporating

Advantages Disadvantages
The regulatory burden is far greater on limited liability companies
than on sole traders and partnerships. A limited company has to
publish annual financial statements. These are public statements of
account, meaning, of course, that anyone (including competitors and
Limited liability reduces the personal
employees) can see how well (or badly) the entity has done. As
investment risk.
stated above, sole traders and partnerships do not have to publish
their financial statements.
Note: the regulatory burden on public limited companies is far
greater again!
It is theoretically more straightforward to
generate funds for a limited company because
new shares can be sold when additional
financing is required. With no cap on the
Limited companies accounts must comply with the relevant domestic
number of different shareholders, investors
or international financial reporting requirements.
could come from anywhere and everywhere.
Indeed, most of the world’s major companies
are public limited companies (plc), which
means their shares are publicly traded.
The existence of a partnership or sole trader is
dependent upon the owners. A limited liability It is a common requirement in domestic law that large companies’
company, on the other hand, has a separate financial statements are audited. In other words, they are subject to
legal identity from its shareholders which an independent review to ensure that they are true and fair, and
means that directors, management and owners comply with all relevant legal requirements and accounting
can come and go but the company will continue standards. This process can be both time-consuming and expensive.
to exist regardless.
Though raising finance is listed as an advantage, note also that
share issues are regulated by law and a public sale of shares on a
There are potential tax advantages. A company
stock exchange can be an extremely expensive affair (and
pays corporate income tax whereas sole
sometimes an expensive failure). It is also quite difficult to reduce the
traders and partners pay personal income tax.
level of share capital or to get rid of shareholders who are deemed to
be standing in the way of management objectives.
Assuming liquid secondary markets, leaving
your position as an owner of share capital in a
limited liability company is far simpler than
exiting your position as a partner.

What is the purpose of a business?


The second question has no clear-cut response. It is commonly believed that a corporate entity’s
objective is ‘to maximize shareholder wealth’. However, this financial objective should be
balanced against a set of non-financial objectives.
Non-financial objectives are typically clustered under three broad sub-headings:

1. ethical;
2. social;
3. environmental.

A business can be a commercial organization, involved with manufacturing, designing,


developing, selling and so forth. Shareholders will likely remain satisfied provided that the risk–
return relationship continues within their accepted bounds. However, there are other stakeholders
who must also be satisfied. It has been argued that ‘the customer is king’, ie the entity exists both
‘for’ and ‘because of’ its customers. The community which the entity serves might be a crucial
factor in relation to performance. Thus, social and environmental policies could dictate relative
levels of success (failure). Investment in plant, property, equipment, employees and so forth is a
function of a business and these investments will, it is hoped, generate jobs and profits which, in
turn, will yield tax revenues which are essential to the economy of the country of residence as a
whole.

The annual report and financial statements


Every business should (and will) produce a summary of their position and performance for a
period of account. There is a set of primary financial statements which is included in the annual
report and financial statements. Their contents and a basic preparation guide will be slowly
introduced to you over the coming pages and chapters.
The IASB reported in the Conceptual Framework for Financial Reporting 2010 (Conceptual
Framework) that ‘the objective of financial statements is to provide information about the
financial position, performance and changes in financial position of an entity that is
useful to a wide range of users in making economic decisions’. The Conceptual Framework
requires any IFRS reporting entity to produce the following information about its economic
resources, claims, and changes in resources and claims:

a Economic resources and claims

Economic resources and claims are recorded in the entity’s statement of financial position.
– Chapter 2 includes a series of exercises and examples which are designed to show you
how to prepare a basic set of financial statements which includes a statement of financial
position and a statement of comprehensive income (income statement). Chapters 3 and 4
provide a summary of how these statements can be used as tools to interpret the
performance and position of the entity.
Economic resources and claims are the entity’s assets and liabilities (see below for
definitions). These can either be current or non-current in nature.
Users need to be aware of the nature and value of an entity’s economic resources and claims
to aid an assessment of the financial strengths and weaknesses, liquidity and solvency
position, and its need and ability to obtain financing [F, OB13].

b Changes in economic resources and claims

An entity should make users aware of, and allow them to distinguish between, any changes
in economic resources and claims which result from either: i) the entity’s performance; or
ii) other events or transactions [F, OB15].
– Changes in the entity’s performance: economic resources might change, for example, as a
result of an increase in production during the period of account. It is likely that at the end of
the year the business will hold more non-current assets (to meet the production
requirements), inventories (unsold units will increase because the size of order will have
comparably increased), trade receivables (ie amounts owed by customers at the end of the
period), trade payables (ie amounts owed to suppliers), and so forth.
– Other events or transactions: a company might, for example, issue new share capital in
order to fund the business. This is, by definition, not a performance-related event.
The changes in an entity’s economic resources and claims resulting from performance are
presented in the statement of comprehensive income [See International Accounting Standard
1 Presentation of Financial Statements {IAS 1}, paragraphs 81–105].
– The preparation of financial statements is covered in Chapter 2 while interpretation of this
financial information is considered in Chapters 3 and 4.
The changes in an entity’s economic resources and claims resulting from other events and
transactions are reported in the statement of changes in equity [See IAS 1,106–110].
Changes in economic resources and claims could impact on users’ assessments of past
performance and a company’s ability to generate future cash flows.

c Financial performance reflected by past cash flows

The statements of financial position, comprehensive income and changes in equity are
prepared on an accruals (accounting) basis. As results reported under accruals accounting
and cash accounting are likely to be different, entities are also asked to prepare a statement
of cash flows [See IAS 7 Statement of Cash Flows].
This statement details how the entity generates cash and spends it [F, OB20].

The elements of financial statements


As we know, the financial statements present information about an entity’s economic resources,
claims, and changes in resources and claims. However, these terms can be confusing because
they are not in everyday use. Instead, it might be more helpful to rephrase these into the
following more commonly known terms (extracted from the Conceptual Framework).
i) Financial position terminology

Asset – a resource controlled by an entity as a result of past events and from which economic
benefits are expected to flow to the entity.

On the face of the statement of financial position these are separated into current (due
within one year) and non-current assets (due in more than one year).
Examples of current assets include inventories, trade receivables and cash held at bank and
in hand. Non-current assets include property, plant and equipment.

Liability – a present obligation of the entity arising from past events, the settlement of which is
expected to result in an outflow from the entity of resources embodying economic benefits.

On the face of the statement of financial position these are separated into current (falling
due within one year) and non-current liabilities (falling due in more than one year).
Examples of current liabilities include trade payables and a bank overdraft. Non-current
liabilities include long-term debt (eg bank loans) and provisions for future costs.

The definitions of an asset and liability contain reference to future economic benefits. This is the
potential to contribute, directly or indirectly, to the flow of cash and cash equivalents to the
entity.

Equity – the residual interest in the assets of the entity after deducting all its liabilities.
Provision – a present obligation which satisfies the rest of the definition of a liability, even if
the amount of the obligation has to be estimated.

ii) Income statement terminology


The IASB has adopted a balance sheet approach to accounting. This means that gains and losses
are measured in terms of changes in assets and liabilities. In other words, an increase in an asset
will give rise to a gain, while an increase in a liability will give rise to a loss. Equally, a decrease
in assets equates to a loss and a decrease in a liability equates to a gain.

Gains – increases in economic benefits.


Losses – decreases in economic benefits.

ILLUSTRATION 1.2 Tesco plc


Tesco plc has recently stood out as one of the UK’s most successful businesses and is one of the world’s largest
retailers. The company employs approximately 500,000 employees in over 6,200 stores and operates across 14
markets.
Expert view 1.6: Complex financial statements
The financial statements presented here are for illustration purposes. Presenting the financial statements of a
fictional entity lacks integrity and risks masking some of the complexities. However, seeing the financial
statements of one of the world’s largest businesses creates issues as well. Do not concern yourself with jargon
you do not recognize but rather embrace the fact that at this stage, you can draw out an elementary
understanding. Once you have completed the financial accounting sections you might want to come back to
this section to see how far your learning has progressed.

Let us work through the introductory pages of their annual report together. It is common for large listed entities such
as Tesco plc to provide an initial summary which brings together the key indicators related to performance and
position related to the current and prior years. Tesco plc have called this section ‘Tesco at a glance 2011/12’ (Figure
1.5).

Figure 1.5 Tesco at a glance 2011/12


If you wish to see this document in full and in colour, follow this link:
http://www.tescoplc.com/files/reports/ar2012/files/pdf/tesco_annual_report_2012.pdf
The annual report typically follows a standardized format which is presented within the first couple of pages (see
Tesco’s contents page in Figure 1.6). This allows you to navigate around what is a sizeable document quickly and
easily. The advent of technology has been particularly useful to search these weighty documents. The annual report
is designed for a variety of user groups and, as such, the portions that are relevant are dependent upon what you
want to discover.

Figure 1.6 Tesco’s contents list


The income statement
Figure 1.7 Tesco’s income statement
Notes

The income statement follows a standard pro-forma.


It is designed to allow a user to assess the performance of the entity over the period of account.
It is only a relatively recent phenomenon that companies are forced to disclose more than their revenue and
profit before tax.
Key line items include: revenue, gross profit, operating profit (otherwise known as profit before interest and tax
[PBIT]) and profit for the year (otherwise known as retained profit).

The statement of financial position

Figure 1.8 Tesco’s statement of financial position


Notes

The statement of financial position is commonly referred to by its previous title, the balance sheet. You will
notice that this balance sheet balances: net assets are £17,801 million and the total equity is also £17,801
million. There is simply a happy coincidence in the terminology however. The balance sheet was named thus
because it is a rephrasing of the expression ‘list of balances’.
It has been said that the statement of financial position is akin to a financial photograph. It is a snapshot of an
entity’s position at a certain point in time capturing summarized details of assets and liabilities.
Though there are several formats permitted under financial reporting regulations, the above format is the most
common in the UK.
The statement is structured in order of liquidity, ie how quickly the asset or liability will be translated into cash.

The statement of cash flows


See Figure 1.9 on the following page.
Figure 1.9 Tesco’s statement of cash flows

Notes

The format of the statement of cash flows is governed by its own financial reporting standard – International
Accounting Standard 7: Statement of Cash Flows (IAS 7).
Cash generated/spent is categorized into one of three sections: operating activities; investing activities; and
financing activities.

Expert view 1.7: The financial year-end


Organizations can set their year-end to whatever date they prefer. There are many determinants, including:
when the financial director/controller is least busy (for example, a toy manufacturer would look to avoid the
busiest times of the year, eg Christmas), when asset values are maximized (eg inversely it is possible that a toy
manufacturer might call their year-end around traditional gift-giving periods when their inventories are most
valuable, eg Christmas), at a time when it is most convenient for your external accountant (possibly reducing
your bill slightly).

Exercises: now attempt Exercise 1.5 on page 40


The articulation of the financial statements
Though this might seem a little confusing to get your head around at first, the following
illustration shows how the financial statements fit together. Feel free to refer back to this as
you’re working through some of the exercises in Chapters 2, 3 and 4.

Figure 1.10 How the financial statements fit together

A brief guide to some key problems and issues with


financial statements
We do not want to finish this chapter on a negative note, but it seems timely to consider some of
the problems with financial statements. You might think that some of these are more important
than others and you’d be right. We will reflect on some of these over the next few chapters but,
for the time being, a summary is sufficient:

Trade-offs between the qualitative characteristics, especially between the enhancing


characteristics. A position might be inherently complex, therefore making it less verifiable,
comparable and understandable.
Owing to the nature of the financial reporting and auditing process, financial statements
generally suffer from not providing timely information. The information contained therein
is based on past transactions and when it is published, it will relate to a period that has
already passed. This is why many argue that the annual report is simply a regulatory
document which carries confirmatory value rather than a positive economic one.
The cost–benefit issue. In other words, at what stage do the reporting requirements become
overly complex and burdensome? Is there a point where the costs of preparation outweigh
the benefits of the information produced?
There is a wide range of users and the potential for conflict between them is vast. The IASB
has promoted the idea that current and potential investors are the primary user group;
however, the annual report and financial statements has not significantly altered to reflect
this. It has been found that the participation of investors in the standard-setting due process
has been extremely limited. Therefore, if an annual report’s first audience is supposed to be
the investor community, then why do they not get more involved?
The conflict between accounting firms’ varied interests means that their participation in the
standard-setting process and in the audit, preparation and presentation of financial
statements leads to potentially schizophrenic behaviour. They must often feel divided
between representing their own self-interest, the public interest, client interests and the
institutional interest.
Standard-setting is a complex process and full agreement is unlikely to be reached given the
conflicting and competing interests of the varied stakeholder groups.
Despite the benefits inherent in the notion, is international accounting an aspiration rather
than an achievable reality?
Selecting appropriate recognition and measurement methods will vary dependent upon the
class of asset (liability) and the information available. For example, derivative financial
instruments (eg forward exchange contracts, options, swaps) cost little to purchase but have
the potential to expose entities to material obligations. Which is a preferable measurement
method? Mark-to-market or historic cost?

Exercises: now attempt Exercise 1.6 on page 41

COMPREHENSION QUESTIONS
1. What are the two main forms of accounting? List the principal differences between the types of information that
practitioners of each discipline produce.
2. What is meant by the terms trade receivables and trade payables?
3. Potential and current investors are the primary users of financial statements but there are others. List four different
users aside from investors and explain their information needs.
4. List the two fundamental qualitative characteristics of financial reporting and provide a brief description of both.
5. List the four enhancing qualitative characteristics of financial reporting and provide a brief description of each.
6. Who are the IASB and what is their role within the IFRS Foundation and standard-setting framework?
7. There are several advantages to incorporating your business, but can you list some of the commonly perceived
disadvantages?
8. Define the three terms: asset; liability; and equity.
Answers on pages 41–44

Exercises
(Answers on pages 44–49)

Exercise 1.1: A comparison between financial and management


accounting
Reflect on the supplementary questions raised in Illustration 1.1. Do you think it is straightforward to ascertain
accurate and reliable answers? Try to write in jargon-free terms an example of the information that might be
prepared by a financial accountant versus a management accountant.

Exercise 1.2: Identifying the users of financial reports


Using the information presented in Figure 1.1, list THREE key user groups for each of the following
organizations and briefly explain what each group might hope to ascertain from the publicly available financial
information prepared by the entity:

(a) Compartmentalized Inc is a company that specializes in making self-fold boxes for storage and
transportation purposes. The entity was founded in 1923. It has grown in a slow but structured and organic
fashion from a family-run business with four staff to an international organization generating revenues of
$85 million with a workforce of over 400 people. The company has no debt and is still family owned.
(b) Gronk plc is a ferry company boasting Europe’s second-largest fleet. The company provides services to
almost every major port on the continent. During the last five years, the company has acquired four
competitors who were in financial distress and facing bankruptcy. This has led to high levels of borrowing.
(c) Worldwide Water is a not-for-profit charitable organization. Their objective is to provide safe drinking water
across Zambia.

Exercise 1.3
Can you provide a suggestion for each of the following, briefly explaining your rationale:

(i) A balance which could be material owing to its magnitude?


(ii) A balance which could be material owing to its nature?
(iii) A balance which could be material owing to its context?

Exercise 1.4
The internationalization of financial reporting has attracted significant attention. There are many in favour and,
equally, many opposed. The United States (US), for example, has continued to use US Generally Accepted
Accounting Principles (GAAP) as opposed to adopting international standards.

(a) Briefly list the key advantages of international financial reporting harmonization.
(b) Briefly outline what you think the key barriers to international reporting adoption might be.

Exercise 1.5
This is a simple exercise to familiarize you with the kind of information that is presented in the financial
statements. Examine the three primary statements presented above to discover the following information:

(a) What is the value of Tesco plc’s property, plant and equipment for the year ended 25 February 2012?
(b) What is the value of their loans and advances to customers for the year ended 25 February 2011?
(c) How much cash and cash equivalents did Tesco plc hold at the end of the most recent financial year?
(d) How much revenue did Tesco plc generate during the year to 25 February 2012?
(e) What was the operating profit for Tesco plc during the years to 25 February 2011 and 2012?
(f) What was the sum of current assets at 25 February 2012? And, as far as possible, can you provide a
breakdown of this balance?
(g) How much cash did Tesco plc generate through financing activities during 25 February 2012?

Exercise 1.6
The IASB acknowledges within the Conceptual Framework that general-purpose financial reports cannot
provide all the information that users may need to make economic decisions. They suggest that users will need
to consider pertinent information from other sources as well.

(a) Briefly outline what you feel the financial statements cannot tell an external user.
(b) List examples of further information (ie beyond the annual report and financial statements) which a
potential investor would require before making an investment (or disinvestment) decision.

Answers to comprehension questions


1. The two forms of accounting are commonly referred to as financial accounting and management
accounting. The differences can be classified into four categories as follows:

Table 1.4

2. Trade receivables arise when a sale has been made on credit but the money that relates to that invoice
has not been collected at the end of the period. This is recorded as a current asset in the statement of
financial position because it is an amount owed to you at the period end.
Trade payables arise when a purchase has been made, the materials/goods received but the payment
has not been made. This needs to be recorded as a current liability in the statement of financial position
as it is an amount owed by you at the period end.
3. Users and their information needs:
Lenders – they have loaned funds to the business under some formal agreement. They require
information as reassurance that the obligations arising as a result of their debt provision will be met.
Suppliers – similar to lenders. They also require some information in order to decide whether to continue
to trade with the business.
Employees – without human capital a business would struggle to survive. Employees are concerned
about the ongoing stability of the firm and the profitability thereof. They might also be interested in
remuneration, especially if bonuses are linked to profitability or financial position.
The public – Businesses have a direct impact on the environment around them, socially, ethically and
environmentally. Therefore, in all these regards, the public might want information regarding a
business’s performance and future prospects.
Government – need information for tax purposes, regulatory purposes, employment prospects, statistics
compilation.
Customers – often have a long-term involvement with the business and they need reassurance regarding
the long-term future. There may be warranty arrangements.
4. The fundamental characteristics:
• Relevant financial information is capable of making a difference in the decisions made by users.
Information may be capable of making a difference in a decision even if some users choose not to take
advantage of it or are already aware of it from other sources. Financial information is capable of making a
difference in decisions if it has predictive value, confirmatory value or both.
• Financial reports represent economic phenomena in words and numbers. To be useful, financial
information must not only represent relevant phenomena, but it must also faithfully represent the
phenomena that it purports to represent. To be a perfectly faithful representation, a depiction would have
three characteristics. It would be complete, neutral and free from error.
5. The enhancing characteristics:
• Users’ decisions involve choosing between alternatives, for example selling or holding an investment, or
investing in one reporting entity or another. Consequently, information about a reporting entity is more
useful if it can be compared with similar information about other entities and with similar information
about the same entity for another period or another date. Comparability is the qualitative characteristic
that enables users to identify and understand similarities in, and differences among, items.
• Verifiability helps assure users that information faithfully represents the economic phenomena it
purports to represent. Verifiability means that different knowledgeable and independent observers could
reach consensus, although not necessarily complete agreement, that a particular depiction is a faithful
representation. Quantified information need not be a single-point estimate to be verifiable. A range of
possible amounts and the related probabilities can also be verified.
• Timeliness means having information available to decision-makers in time to be capable of influencing
their decisions. Generally, the older the information is the less useful it is. However, some information
may continue to be timely long after the end of a reporting period because, for example, some users may
need to identify and assess trends.
• Financial reports are prepared for users who have a reasonable knowledge of business and economic
activities and who review and analyse the information diligently. Classifying, characterizing and
presenting information clearly and concisely make it understandable.
6. The International Accounting Standards Board – or IASB – acts as an independent standard-setting body
on behalf of the IFRS Foundation. The IASB currently has 15 full-time members who are responsible for
the development and publication of IFRSs. They actively engage with stakeholders from around the
world, including investors, analysts, regulators, business leaders, other accounting standard-setters and
the accountancy profession. Their principal objective is to develop a single set of high-quality,
understandable, enforceable and globally accepted international financial reporting standards (IFRSs).
7. The disadvantages of incorporation include the following:
(a) The regulatory burden is far greater on limited liability companies than on sole traders and
partnerships. A limited company has to publish annual financial statements. These are public
statement of account, meaning, of course, that anyone (including competitors and employees) can
see how well (or badly) the entity has done. As stated above, sole traders and partnerships do not
have to publish their financial statements. Note: the regulatory burden on public limited companies is
far greater again!
(b) Limited companies accounts must comply with the relevant domestic or international financial
reporting requirements.
(c) It is a common requirement in domestic law that large companies’ financial statements are audited. In
other words, they are subject to an independent review to ensure that they are true and fair, and
comply with all relevant legal requirements and accounting standards. This process can be both
time-consuming and expensive.
(d) Though raising finance is listed as an advantage, note also that share issues are regulated by law
and a public sale of shares on a stock exchange can be an extremely expensive affair (and
sometimes an expensive failure). It is also quite difficult to reduce the level of share capital or to get
rid of shareholders who are deemed to be standing in the way of management objectives.
8. Definitions of some key financial terms:
(a) asset – a resource controlled by an entity as a result of past events and from which economic
benefits are expected to flow to the entity;
(b) liability – a present obligation of the entity arising from past events, the settlement of which is
expected to result in an outflow from the entity of resources embodying economic benefits;
(c) equity – the residual interest in the assets of the entity after deducting all its liabilities.

Answers to exercises

Exercise 1.1 A comparison between financial and management


accounting
The preparation of the annual report and financial statements – statement of financial position, statement of
comprehensive income, statement of cash flows – is the domain of the financial accountant. Though the
information is relevant to the senior management team and board of directors, one would hope that this annual
exercise was an opportunity to review and bring together rather than learn. It is principally driven by an external
demand and therefore is largely externally focused. The financial statements are prepared according to rules
and regulations. They are prepared at certain points in time, normally annually. They are prepared using
historic information, are based upon records of past transactions and therefore have the tendency to be
backwards-looking.
The management accountant would be concerned with fundamental problems such as setting prices and
determining costs. From this, the management accountant would seek to provide information that would help
the senior management team. For example, through setting budgets, evaluating the working capital position
(current assets versus current liabilities), appraising investment opportunities and guiding strategic decisions.

Exercise 1.2 Identifying the users of financial reports


(a) Compartmentalized Inc:

Owners: This is a family-run company and therefore the key information users will be the family
themselves. The business is now a large organization and therefore accurate record-keeping is
essential as a management discipline. Often the preparation of financial statements is in itself a
worthwhile venture as it ensures that at least once a year, the various parts of the business are brought
together, accounts are reconciled and balances examined. They might use the information to guide
their decision making. If they prepare segmental information then it might be an opportunity to focus on
relative levels of profitability between business areas.

Employees: The financial statements often add some level of transparency to employees. If the business
has performed satisfactorily then any pre-existing feelings of instability, insecurity or unrest might be
quelled. If staff bonuses are linked to performance or position then the financial statements might shed
some light on the level of remuneration afforded.

Competitors: Competitors could use Compartmentalized Inc’s financial information to reflect on their own
performance and position. This could be used for many decisions, including pricing, product cost,
outsourcing and financial strategy.

(b) Gronk plc

Current and potential shareholders: Gronk is a listed company and therefore the management has
stewardship responsibilities to their current shareholders. Their position and performance will influence
their current investors’ view on whether to retain their level of investment, increase it or reduce it.
Potential investors will also be advised whether to buy stock in Gronk. Normally this advice will be
provided by investment analysts and the instructions will be conveyed through institutional investors.
Information contained within the annual report is used to inform this advice and to prepare valuation
estimates.

Lenders: Gronk plc has significant borrowings. The lenders will consult the financial information to ensure
there have been no loan covenant breaches. They might also use the information to gauge the level of
risk of current and future lending.

Customers: Customers will often book their holidays months in advance of their departure date. If
appropriate (insurance) cover is not taken then customers leave themselves exposed to entity-specific
risk. Assuming the financial information released is positive, then customer doubts and fears could be
assuaged.

(c) Worldwide Water

Public – charitable donors: Funds are accumulated through the goodwill of the public who make donations.
The public, therefore, would be interested to see how effectively and efficiently the charity is using their
financial resource. They might be looking to reappraise their charitable giving and therefore they could
attempt to perform a reconciliation of the amount of ‘good’ done for every dollar spent.

Public – charitable donation beneficiaries and their advocates: It is possible that the beneficiaries, and
those who act on their behalf, will also want to see how effectively the resources of the charity have
been managed.

The government and her regulatory bodies: Though taxation is not a concern as Worldwide Water is a not-
for-profit entity, the government might provide financial (and non-financial) support. It is likely, therefore,
that the charity will have some form of financial accountability and associated regulation. Often the
government will supply charities with non-financial targets which might drive financial rewards.

Exercise 1.3
There are many possible alternatives. The following are among them:

(i) A balance which could be material owing to its magnitude? The problem with identifying balances
according to their magnitude (or size) is that the value will differ by entity. There are some rules of thumb
that can be employed. Some organizations use 5 per cent of profit before interest and tax to determine
what is material from what is not. However, for an entity whose worth is high but profits are low, this can
cause significant frictions and drive an inappropriate level of focusing on detail which is not relevant to a
user’s understanding of the performance or position of the entity (ie immaterial). Some organizations use a
combination of income statement measure and financial position measure, eg 1 per cent of net assets + 1
per cent of turnover.
Balances which are material by their magnitude will also differ dependent upon the industry and
financial strategy of the entity. For example, a manufacturing firm is likely to have high inventories and
therefore these will probably be material owing to their magnitude. An airline, however, which is likely to
have low inventories – they will stock on-board food, drinks and small amounts of product for in-flight duty-
free sales which will be a fraction of the value of the aeroplanes and corresponding liabilities – is unlikely to
need to focus their energy on their inventories balance.
Misstatements within individual balances might be material owing to their context (see iii).
(ii) A balance which could be material owing to its nature? There are several examples of balances which are
material simply because of their nature. For example, any misstatement of an entity’s issued share capital
is likely to be deemed material. Equally, there are some balances which are deemed material because
they are sensitive, for example directors’ remuneration.

(iii) A balance which could be material owing to its context? An error of $20,000 in a revenue balance of $1
million is unlikely to be material. However, if a customer files for bankruptcy owing you $20,000 and your
total trade receivables are $100,000 then this sum would be considered to be material in context.

Exercise 1.4
The key advantages of international financial reporting harmonization include the following:

comparability;
information access;
aids internal communication;
makes the appraisal of foreign entities more straightforward;
benefits to global accounting firms and their accounting staff;
ease of transfer of staff.

And so on…

The key barriers to international reporting adoption:

different purposes for preparing financial information;


different legal systems;
different user groups;
nationalism.

And so on…

Exercise 1.5
(a) £25,710 million
(b) Non-current portion: £2,514 million; current portion: £2,657 million; total: £5,171 million. (Note: prior-year
figures are provided alongside current year balances.)
(c) £2,305 million. (Note: you can extract this balance from either the statement of financial position or the
statement of cash flows; by default they must agree.)
(d) £64,539 million.
(e) 2011: £3,917 million; 2012: £3,985 million.
(f) £12,863 million, broken down as shown in Table 1.5.

Table 1.5 Exercise 1.5(f)

£m
Inventories 3,598
Trade and other receivables 2,657
Loans and advances to customers 2,502
Derivative financial instruments 41
Current tax assets 7
Short-term investments 1,243
Cash and cash equivalents 2,305
Assets of the disposal group and non-current assets classified as held for sale 510
Total 12,863

(g) £(1,366) million. (Note: the figure is negative, ie cash outflow. This is largely because the group paid out
£1,180 of dividends to equity owners.)

Exercise 1.6
(a) Brief outline of what the financial statements cannot tell an external user:
Internally, one would hope that all relevant financial information related to the entity would be freely
available between the senior management team. However, external users do not get to see the same level
of detail related to account balances. Issues such as individual customer accounts and exposure to
creditworthiness are not freely disclosed. A company might show certain assets at an over- or under-value
owing to an accounting policy choice, but as an externality you would not be aware of such things. The
information being passed around internally is up to date and often forecasted ahead, taking into account
project-specific risks. The information an external user has available often lacks this timeliness and any
adjustment for risk tends to be based upon assumptions rather than facts.
The financial statements are audited but the front end of the annual report is simply checked for
consistency. This might mean that narrative and graphical representations can lead to differing external
interpretations, and sometimes misinterpretations. Not only does the potential for impression management
exist within the textual and graphical narratives, there is also the opportunity for the entity to engage in
earnings management (sometimes referred to as smoothing) practices. Indeed, the preparation of financial
statements often requires managers use their judgement because of choices in accounting policy. Among
some accountants, the employment of managerial judgement is both welcomed and supported because it
can lead to more meaningful – and hence useful – information.
(b) Examples of further information required before making an investment (or disinvestment) decision:
There was a time when the release of the annual report and financial statements was the most important
moment in the financial calendar for followers of a firm. Though it is still significant, the acceleration in
sophistication of communication networks and the media has meant that this level of importance has
diminished. There are many examples of information not contained within the annual report. The role of
investment analysts is now ultra-competitive and the relative ability to gather public and private information
from disparate and far-ranging sources might make the difference to one’s career.
Retail analysts, for example, can go to absurd lengths. It is not a simple case of reading the broadsheet
newspaper headlines or following the entity’s press releases. They could investigate potential next
opportunities for new markets by reading trade journals, visiting possible sites, speaking to local people,
attending council meetings and so forth. The new constraints on analysts are time and resource, rather
than lack of timely information or sophistication in interpretation.
02
Accounting concepts and systems

OBJECTIVE
We use this chapter to introduce the types of financial accounting information you will encounter in your daily managerial
lives. We provide a basic guide to bookkeeping and outline how to prepare a straightforward set of financial statements.
Some measurement rules and accounting concepts are also explained.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Describe the key elements of an accounting system.


Identify the different types of accounting adjustment.
Assess the impact of accounting adjustments on reported results.
Prepare basic primary financial statements.

KEY TOPICS COVERED


What accounting systems record and how.
Cash-based accounting, accruals and other accounting adjustments.
The impact of accounting adjustments on the statement of financial position and on profit.
Practical issues of adjustment, for example dealing with tangible non-current assets.

MANAGEMENT ISSUES
This chapter will deal specifically with the management skills of understanding and assessing the impact of accounting
adjustments on the reported results.

Introduction
During this chapter we try to show you why financial statements are important, describe their
contents and show you how to prepare a straightforward set. Appendix A deals with double-entry
bookkeeping and works through the same examples using this method.
Even though this material represents only the tip of the iceberg in terms of depth and breadth,
it is extremely helpful to know how information emerges out of data. In turn, the series of
exercises contained herein should show you that the quality, complexity and context are crucial
to grasp before you go on to try to form an assessment and appraisal of an entity’s position and
performance.
This chapter also introduces some key accounting concepts and conventions. These should
reveal that interpreting – and to an extent, preparing – financial statements is an art, not a
science. This also highlights how accounting procedures and policies can impact on reported
results. As managers, you need to be aware that accounting is not a system of rigid or fixed rules
where output is predefined. Rather, there is flexibility within the regulations which allow scope
for presentational, recognition, measurement and disclosure differences.

What is the purpose of the financial statements?


This question is the most obvious starting point for this chapter. Towards the end of the previous
chapter we showed you Tesco’s financial statements. We hope that you have also gone away and
looked up a couple of companies’ annual reports in which you have a special interest. For Tesco,
we showed you three primary financial statements: the statement of comprehensive income
(income statement); the statement of financial position; and the statement of cash flows. This
chapter will focus on these and walk you through the accounting for some basic transactions.

Statement of comprehensive income (income statement)


The income statement shows how much profit the business has made during a specific period of
time. We told you that the ‘period of account’ is normally a year, but large organizations are
often required to produce interim statements, eg quarterly.
The layout of the income statement will depend on the entity, but there are a selection of
headings which are typical, including:

Gross profit: A company generates revenue by undertaking its normal operating activity,
eg selling goods and/or services. The costs that directly relate to selling these goods (‘cost
of goods sold’; or more usually, ‘cost of sales’) is deducted from the revenue. This
generates the gross profit figure.
The gross profit is stated before the deduction of indirect expenses – such as operating
expenses, finance costs or tax – and before adding any ‘other’ income generated, not arising
from the normal operating activity – such as finance income.
Operating profit: Business expenses which are not directly incurred in relation to the
generation of revenue are deducted after the gross profit line. These indirect costs might
include items such as rental costs, business rates, heating and lighting costs, administrative
expenses, wages and salaries, depreciation and so forth.
Finance income and/or costs (which include interest receivable/payable) and taxation are
deducted from the operating profit, which is why the operating profit is sometimes referred
to as profit before interest and tax (PBIT) or earnings before interest and tax (EBIT).
Profit for the year (period): The profit for the year is arrived at by deducting all other
expenses from the operating profit. This total is transferred to the statement of financial
position on a rolling basis each year. The balance on this account accumulates under the
heading ‘Retained earnings’.

Figure 2.1 shows an income statement for a fictional entity – Malambo Inc. – for illustration
purposes.

Figure 2.1 Malambo Inc

We recommend that you search on the internet for your ‘most (least) favoured brand’ or ‘most (least) favoured
company’. Sometimes, a personal involvement or an ability to contextualize makes a subject more meaningful.

Expert view 2.1: Comprehensive income


The statement of comprehensive income is actually subdivided into two parts. The top part is the ‘income
statement’ and the bottom part is the section that captures the ‘comprehensive income’. As a basic rule of
thumb, the income statement shows an entity’s realized gains and losses whereas the statement of
comprehensive income as a whole identifies those which are both realized and unrealized. Examples of
unrealized balances could include gains and losses arising from issues such as revaluing property to its open
market value, cash-flow hedging gains and losses, and foreign exchange translation gains and losses.

Online content
Listed companies such as Tesco plc have higher levels of reporting requirements and are subject to greater
levels of scrutiny and regulation. We have provided a couple of examples (available online) which show that
there is still freedom to present information differently.

The statement of financial position (balance sheet)


The statement of financial position (commonly referred to as the balance sheet) is akin to a
financial photograph: a snapshot of the current worth of an organization at one moment in time.
Financial statements are prepared according to one golden rule, as follows:

Every debit entry in the ledger accounts must be matched by a corresponding credit entry.

And nowhere is this rule more visible or understandable than in the statement of financial
position. The accounting equation is as follows:

Assets – Liabilities = Capital (shareholders’ funds).

Both assets and liabilities are categorized as either current (due within one year) or non-current
(due in more than one year). Below is a brief description of each category and a few examples:

Non-current assets are bought with the intention to use them to generate revenue over
the course of a number of years. There are two forms of non-current assets – tangible and
intangible. The former are those which it is possible to touch, see and feel. Intangible assets
are assets that have a realizable market value despite not having a physical presence. The
rules governing the capitalization (ie bringing onto the statement of financial position as an
asset) of intangible assets are quite strict. Analysts typically reassess the value of an entity’s
intangible assets, as they are commonly a key driver of the difference between market value
(ie the share price) and book value (ie the notion of net worth shown at the foot of the
statement of financial position).
– Examples of tangible non-current assets include: property (land and buildings), plant,
equipment, fixtures, fittings, motor vehicles and so forth.
– Examples of intangible non-current assets include: (purchased) goodwill, patents,
royalties, computer software and so forth.
Current assets are expected to be sold or converted into another form within one year.
– Examples include: inventories, trade receivables, cash held in hand or at bank, cash
equivalents, prepayments.
Non-current liabilities are amounts owed by the business which it is not obligated to
repay within the next 12 months.
– Examples include: bank loans, mortgages, debentures, loan stock, provisions and so forth.
Current liabilities are obligations which need to be settled within the next 12 months.
– Examples include: bank overdraft, short-term borrowings, trade payables, taxation and so
forth.

Online content
We have provided a couple of examples (available online at: www.koganpage.com/accountingfm) which
show some complex entities’ statements of financial position to mirror the income statements provided. As
before, we recommend that you search for brands and companies you are acquainted with and have a look at
their annual reports.

The statement of cash flows


The statement of cash flows shows whether we have earned or spent cash (and cash equivalents)
during the year. The focus of the statement of cash flows is, as the name suggests, cash inflows
versus cash outflows. The statement is divided between three activities: operating; investing; and
financing.

Expert view 2.2: Preparing a statement of cash flows


Students of financial accounting commonly find this the most difficult of the three primary statements to
prepare. This is surprising because you often know the answer before you start – if you know the opening and
closing cash position and (available from the statement of financial position or from your books and records),
then preparing the statement of cash flows is simply a reconciliation exercise where you organize movements
in balances in to appropriate categories (ie decide whether they are operating, investing or financing).

Note that the income statement and balance sheet are prepared on an accruals basis whereas
this statement is prepared on a cash basis. This is important because organizations can be
profitable (or loss-making) but be losing (or gaining) cash. We will explain this in more detail
later. Ultimately, the accounting profits and the cash profits will reconcile, but the chances of
them being identical year to year are negligible.
The key differences between profits on an accruals basis versus those measured on a cash
basis relate to:
Timing differences. For example, goods might be sold to an entity, delivered
immediately and an invoice for payment raised at the same time. However, if you grant a
credit period then you should expect your customers to take advantage of it (after all, this is
essentially an interest-free loan for them). Therefore, the revenue can be booked to the
income statement as the transaction took place during the period, but if the cash has not
been received by the end of the period then it will show as a balance being owed to you. In
other words, the cash follows the sale.
Accounting estimates. It is necessary for accountants to make accounting estimates
during the preparation of the financial statements. One example is the accounting for
depreciation against tangible non-current assets. Depreciation acts as a proxy for the costs
related to wear and tear of an asset over time, so assets classified as non-current must be
written off to the income statement as their revenue-generating potential is consumed (more
on this later). In cash terms, however, the payment will often occur up-front.
Accounting transactions that bypass the income statement. For example, capital
expenditure, sale of share capital, repayments of loans and so forth.
Changes in the working capital position. For example, an increase in inventories
means that an entity has invested cash in inventories, which it hopes to translate back into
cash through the sale of these goods. Figure 2.2 shows the working capital cycle and reveals
how the process of investment and re-investment in working capital makes the management
of short-term resources crucial. It also shows the importance to a business of focusing on
the cash cycle.

Figure 2.2 Working capital cycle

Expert view 2.3: Financial reporting regulation


Financial reporting is governed by many regulations, foremost of which are the International Financial
Reporting Standards (IFRS). The predecessor to the IASB was called the International Accounting Standards
Committee (IASC). This body released International Accounting Standards (IAS). Though IAS’ are in a process
of being developed and refined, it is intended that they will all be replaced by IFRS’. This exercise is far from
being complete and many still exist. Therefore, you will commonly see references to IAS as well as IFRS. For a
full list see Appendix B.
We will come back to this idea again in later chapters but, for the time being, it is worth noting
that the concept of cash is important in both financial and management accounting. Cash is often
described as the ‘lifeblood’ of the business because without cash, a business cannot expect to
survive for long.
Figure 2.2 illustrates how a traditional manufacturing business operates while also serving to
highlight the importance of cash to the process. While a company holds inventories or they are
owed money, the cash balance is reduced. This can be offset by owing others money (payables),
but if management extend their credit terms too far, beware the consequences. This is, of course,
an over-simplification but it does highlight the importance of efficiency.
When preparing a statement of cash flows, you will notice that an increase in a current asset
(eg inventories) will lead to a decrease in cash. We hope that Figure 2.2 makes it obvious to see
why.

Exercises: now attempt Exercise 2.1 on page 85

Preparing a set of financial statements


Even though it may never be necessary for you to be called upon to prepare a full and detailed
set of financial statements, an understanding of the basics is essential to any interpretative study
of financial information. During this section, our aim is to cover only the basic, but most
valuable, tenets.

Double-entry bookkeeping: an introduction


Double-entry bookkeeping is a system of recording financial transactions. The underpinning
principle is that every debit entry must have an equal and opposite credit entry. Recognizing
amounts in a timely and structured manner has helped owners and managers organize their day-
to-day business for millennia. Today, there are many computerized bookkeeping packages which
are extremely useful in assisting accounting staff to produce key reports. These accounting
packages, however, do not teach you accounting per se; they just simplify the process of data
entry.

Expert view 2.4: Mastering the basics of financial accounting


Financial accounting should be seen as a building-block exercise. You should strive to master the skills and
knowledge gathered at each stage before moving on to the next. There is one caveat, and this relates to
double-entry bookkeeping. Many accountants talk of a ‘eureka’ or a ‘light bulb’ moment. What they mean is that
one day they woke up and found that double-entry bookkeeping simply makes sense! Thinking logically always
helps. Every time you engage in a financial transaction, imagine the double-entry bookkeeping system at work.
For example, you buy a newspaper from the shop. You have exchanged your money for an asset of (deemed)
equivalent worth.

WORKED EXERCISE 2.1 Mobius (1)


Let us suppose that you wish to start your own business, Mobius Inc. The following transactions shown in examples
Mobius (1) to Mobius (3) represent the first week of trading for the entity.
On day 1, you opt to put financial distance between you and the trading entity and transfer $1,000 from your
personal bank account to a bank account you hold in the name of the new enterprise – Mobius Inc.
Assuming that the new business is a separate legal entity, how does this affect the accounting equation?

NOTE: As you can see, the accounting equation captures both sides of the transaction and only by making the entry
twice does the equation (and, by extension, the statement of financial position) balance.

Continuing this exercise…


On day 2, you borrow $500 from a friend to provide further financial help to your business. How does this impact on
the accounting equation?

NOTE: The effect of the financial transaction is to increase assets by $500 as the amount borrowed from your friend
would be deposited in the current account of the business. The liabilities would increase by $500. This represents the
new loan the business has taken on. The capital that you invested has neither increased nor diminished in the
process of this transaction and therefore remains at $1,000. This, of course, means that your accounting equation
continues to balance.

The transactions above illustrate that if an adjustment is made to one side of the equation, you must make an
identical adjustment either to the other side of the equation or to the same side.

Exercise 2.2: Accounting equation adjustments


Continuing the worked exercise above, make the necessary adjustments for the following transactions:

(a) On day 3, Mobius Inc invests $500 of cash by acquiring a new computer.
(b) On day 4, Mobius Inc buys some raw materials worth $400 and holds them as inventories. The cash
required to settle the invoices related to these purchases does not need to be found for 10 days as these
are the credit terms offered.

The transactions above relate exclusively to the statement of financial position. They require
making changes to assets, liabilities and equity with no profit implications. This is overly
simplistic, therefore these next worked exercises of individual transactions provide the
opportunity to see how the accounting equation changes when an entity trades.
WORKED EXERCISE 2.2 Mobius (2) – trading activities
On day 5, Mobius Inc uses the raw materials purchased the previous day to produce 30 units of finished goods stock.
On day 6, half of these are sold for $50 per unit. Cash is received immediately for five of those sold. The remainder
were sold to customers on 10-day credit terms.

Solution
The transfer between raw materials and finished goods inventories has no impact on the accounting equation (in
reality it is difficult to imagine that no value is added to the product [eg extra materials, direct labour], but for
simplicity’s sake, let us assume that this is the case). We would simply transfer $400 from one category of current
assets to another – raw materials inventory to finished goods inventory. Therefore, at this stage, the accounting
equation remains unchanged, as follows:

However, the sale of the product does generate a change in the entity’s net worth. Ultimately, most organizations hold
as their corporate objective the maximization of shareholder wealth and it is unsurprising that goods are bought and
subsequently every effort is made to sell them for a profit. In this case, the goods cost $400 and were transformed
into 30 saleable units of inventory, of which half were sold for $50 per unit. Five were paid for immediately while 10
have been sold under 10-day credit terms. Therefore, the changes to the accounting equation that we will record are
as follows (see Table 2.1):

(i) Revenue generated: 15 units * $50 = $750


(ii) Of which, cash collected was 5 * $50 = $250 and amounts owed from customers was 10 units * $50 = $500
(iii) Amount of inventory used to generate these sales was $200.

Table 2.1

Assets $ Liabilities $ Capital $


Opening balance 1,900 Opening balance 900 Opening balance 1,000
i) Revenue 750
ii) Cash collected 250
ii) Owed by customers 500
iii) Inventories consumed (200) iii) Cost of goods sold (200)
Closing balance 2,450 900 1,550

The effect on assets and liabilities is straightforward. Assets increase due to the extra amount of cash held and the
amount of money owed by customers. Liabilities remain unchanged.
Why has the level of shareholder’s funds changed? The business has managed to generate $550 of profit ($750 of
revenue less $200 of inventories consumed). These retained profits are added to the shareholder’s funds and remain
there until spent or distributed.

WORKED EXERCISE 2.3Mobius (3) – the financial statements [week 1]


Preparing complex financial statements at this stage of your accounting learning is not necessary. However, it is
possible to translate the information we have into a basic set of summarized financial statements:
* Transfer to statement of financial position at the end of the period of account

NOTE: transaction numbers are shown in square brackets to help identify the matching accounting entry. You will
notice that after each transaction is entered, the statement of financial position could be closed off and would
balance, ie the top half would equal the bottom half.

Expert view 2.5: Revenue and expenses


The IASB would prefer we thought of gains and losses arising as a result of assets and liabilities increasing and
decreasing. However, it is probably more common for non-accountants to think of profits as being generated
when revenue exceeds expenses.
Revenue typically relates to the income an organization generates by selling goods and services.
Expenses (costs) are incurred in the process of generating revenue.

At the end of the period of account, an income statement is reset to zero because this statement reports ‘for the
period ended’, whereas the statement of financial position shows the position of the entity ‘as at the period ended’.
Therefore, let us assume that the above transactions represent Mobius Inc’s position as at the end of the first week
of trading and the business is about to begin the second week. The income statement balances would be reset to
zero and the profit of $550 taken to the statement of financial position where it will be retained until it is used
(consumed by losses or distributed).

The income statement: cost of sales working


The cost of sales working in the statement of financial position applies the accruals concept to
match the expense to the period in which it is incurred. The cost of sales is sometimes referred to
as ‘the cost of goods sold’. In other words, we are matching the units of product sold to the direct
cost of purchasing (producing) those units. The basic working is as follows:

Cost of sales
Opening inventories x
Purchases x
(Closing inventories) (x)
x

Exercises: now attempt Exercise 2.3 on page 86

WORKED EXERCISE 2.4 Mobius (4) – the financial statements continued


Now that we have a statement of financial position as at the end of week 1, we can adjust straight to the face of the
statements and thus monitor how the business’s net worth and profitability levels change during the second week of
trading, as shown in Table 2.2.

Table 2.2

Day of trading

The business acquired raw materials valued at $800.


8
Invoices need to be paid within 10 days.

9 All the raw materials were converted into 60 units of finished goods inventories.

The company received a utilities bill for $200 which it paid immediately out of cash.
10
Note: assume this charge relates exclusively to this period of account.

Mobius Inc. sold 35 units of inventories for $60 per unit. Cash was received for 20 of these units and the
11
remainder were offered to customers on 10-day credit terms.

Mobius Inc. bought a printer/scanner for $100.


12
Mobius Inc. bought some stationery including paper and envelopes for $50.
13 Mobius Inc. received $500 from customers who had bought goods on credit terms.

Amounts invoiced by suppliers during week 1 of $400 were paid.


14
Mobius Inc. paid a web designer $750 to develop an online presence for the business.

Remember: bookkeeping is a dual-aspect method and you will need to post every transaction twice for the
statement of financial position to balance.

Solution
The exercise has been completed in order of appearance in the text above. All transactions have been accurately and
appropriately double-entered and can be traced through by reference to square brackets. The opening balances from
week 1 have simply been brought forward and therefore have no transaction number.

Mobius Inc
Summarized Statement of Financial Position
As at the end of week 2
Assets
Non-current assets
Computer (500) 500.00
Printer/scanner (100 [5]) 100.00
Current assets
Cash at bank (1,250 – 200 [3] + 1,200 [4] – 100 [5] – 50 [6] + 500 [7] – 400 [8] – 750 [9]) 1,450.00
Inventories
Raw materials (0 + 800 [1] – 800 [2]) 0.00
Finished goods (200 + 800 [2] – 466.67 [4]) 533.33
Trade receivables – amounts owed by customers (500 + 900 [4] – 500 [7]) 900.00
Liabilities
Loan from friend (500) (500.00)
Trade payables – amounts owed to suppliers (400 + 800 [1] – 400 [8]) (800.00)
Net assets 2,183.33
Shareholder’s funds (capital and reserves)
Capital (1,000) 1,000.00
Retained profits
Brought forward 550.00
Profit generated during week 2 633.33
Shareholder’s funds 2,183.33
Mobius Inc
Summarized Income Statement
For the period ended DD/MM/YYYY (week 2)
Revenue (2,100 [4]) 2,100.00
Cost of sales (466.67 [4]) (466.67)
Gross profit 1,633.33
Expenses:
Utilities (200 [3]) (200.00)
Printing, postage and stationery (50 [6]) (50.00)
Web development costs (750 [9]) (750.00)
Profit 633.33

Exercises: now attempt Exercise 2.4 on page 87

Underlying concepts: measurement rules and


fundamental accounting concepts
Let’s take a short break from these preparation exercises and turn our attention to some
measurement rules and underlying concepts. You’ll appreciate that not everything is as easy to
account for as, say, exchanging cash for a newspaper. Indeed, the above exercises portray a
series of simple transactions during which we have deliberately minimized your exposure to
recognition, measurement and presentation problems. The following section outlines some of the
potential complexities and the guidelines accountants have developed to deal with them.
Following this, we will pick up the case of Mobius Inc again and work through some more
preparation exercises to illustrate how these concepts are articulated through numbers rather than
text.

Measurement rules
There are some basic measurement rules which are applied in financial accounting. These rules
explain how balances are recorded in the financial statements. Owing to the importance of the
measurement basis used for the values recorded in the financial statements, it is unsurprising that
few areas have attracted so much academic and professional commentary. The following are
considered to be key:

historical cost accounting;


current cost accounting;
mixed measurement model;
money measurement concept;
business entity concept.

Online content
More information is provided about these measurement rules online at: www.koganpage.com/accountingfm

Fundamental accounting concepts


There are a number of fundamental accounting concepts which are enforced either through
regulation or legislation. The following concepts are relevant:

accruals concept;
going concern;
prudence;
disaggregation;
materiality.

A brief review of each follows.

Accruals concept
The accruals concept is often referred to as the matching concept or matching rule. Indeed, the
reference to ‘matching’ is a simpler way to envisage its operation (even if it does not strictly hold
in some domestic GAAP). The accruals concept states that revenues, profits and the associated
costs should be matched to the same period’s income statement.
There are many everyday instances where a transaction would be recorded in different
accounting periods dependent upon whether you employed an accruals accounting system or a
cash accounting one. For example, electricity bills are commonly issued and paid in arrears.
Imagine that you have 11 monthly invoices related to your electricity costs for the year ending 31
December and are awaiting the 12th and final invoice. You will need to accrue for the expense
because the electricity was being consumed during the period of account. You will have to bring
the cost into this period’s income statement and provide for it by setting up a corresponding
liability in the statement of financial position.
Extending this example, it is possible that you might need to estimate the level of
consumption for the 12th month because the electricity company hasn’t issued their invoice
before your accounts finalization date. It is easy to see how accruals accounting can lead to
increased subjectivity because sometimes it is necessary to estimate the value of future
transactions.

Expert view 2.6 Do not confuse accruals and the accruals concept
Though the terms are clearly related, you should be careful not to define an accrual in the same way as you
define the accruals concept. An entity would need to accrue for the electricity cost (ie bring the cost into this
period’s income statement) by setting up an accrual (ie set up a current liability in the statement of financial
position) because of the accruals concept (ie matching concept). The accruals concept is also responsible for
giving rise to assets (ie prepayments) and associated gains.

Going concern
Entities are required to prepare accounts on the basis that they are a going concern. This means
that the business is commercially viable, able to pay its obligations as they fall due, and whose
owners (or other controllers) intend it to continue in operation for the foreseeable future. In
particular, when an entity provides assurances over its going concern status, one should be able
to assume that the entity will not go into liquidation or scale down its operations in a material
way within a period of 12 months or more.
Despite standard-setters and professional bodies assuring investors that they should not panic
if an entity records a statement declaring that they are not a going concern, one should note that
it is extremely difficult to find an annual report containing this statement. Providing a statement
that an entity is not a going concern, some might argue, should be interpreted as a signal of
transparency and openness. Instead of presenting an explicit negative statement about their going
concern status, companies facing an uncertain future tend to refer to doubts over their foreseeable
future.
The key financial accounting complication of not maintaining a going concern status is that
the assets and liabilities should be valued and shown at their ‘break-up’ value ie the amount they
would sell for if they were sold off piecemeal and the business were broken up. For example,
non-current assets would need to be presented as current assets. Their previously recorded value
(ie cost less accumulated depreciation) would need to be adjusted to their ‘forced sale’ value.

Expert view 2.7: Going concern (extracts of going concern disclosures)


Though there are some jurisdictions (eg the United States) where a statement of an entity’s going concern
status is not expressly required, the directors of most companies include some reference to the likelihood of
continuing to trade as normal into the foreseeable future. Indeed, such a statement is roundly encouraged by
stakeholders even where not required (as shown by Gallagher and Paul, 2012). Examples from two companies
follow: first, Whitbread plc, a major retail brand conglomerate; and second, Oxford Instruments plc, a leading
provider of high-technology tools and systems for research and industry.

i) Whitbread plc
Annual report and accounts 2011/12 (p 27)
Available from Whitbread plc’s download centre: www.whitbread.co.uk/global/download-centre/reports-
and-presentations.html.

Going concern
A combination of the strong operating cash flows generated by the business and the significant headroom on
its credit facilities support the directors’ view that the Group has sufficient funds available for it to meet its
foreseeable working capital requirements. The directors have concluded that the going concern basis remains
appropriate.

ii) Oxford Instruments plc


Annual report and financial statements 2012 (p 21)
Available from Oxford Instrument plc’s download centre: www.oxford-instruments.com/investors.

Going concern
The Group’s business activities, together with the factors likely to affect its future development, performance
and position, are set out in the chief executive’s statement. The financial position of the Group, its cash flows,
liquidity position and borrowing facilities are described in this financial review.
The diverse nature of the Group combined with its current financial strength provides a solid foundation for a
sustainable business. The directors have reviewed the Group’s forecasts and considered a number of potential
scenarios relating to changes in trading performance. The directors believe that the Group will be able to
operate within its existing debt facility which expires in December 2014. This review also considered hedging
arrangements in place. As a consequence, the directors believe that the Group is well placed to manage its
business risks successfully.
The financial statements have been prepared on a going concern basis, based on the directors’ opinion,
after making reasonable enquiries, that the Group has adequate resources to continue in operational existence
for the foreseeable future.

Neutrality (and/or) prudence


Accounting transactions and other events are sometimes uncertain and yet, in order for the
information to be useful for decision making and fulfil the fundamental qualitative
characteristics, these uncertain transactions and events still have to be reported in a timely
manner (to correspond to the appropriate period of account). Thus, it is necessary for
management to make estimates that counteract the uncertainty. Historically, it has been
preferable to err towards prudence. This is not to say that accuracy is not important to
accountants and the information which they produce.
The European Financial Reporting Advisory Group (EFRAG), alongside the national
standard-setting bodies of France, Germany, Italy and the UK, published a joint bulletin in April
2013 outlining their position on the prudence convention. Within this bulletin they drew attention
to the history of prudence and its usefulness. The Fourth EU Directive on Company Law of 1978
requires that ‘valuation must be made on a prudent basis’ and that, in particular, only profits
made at the balance sheet date may be included, whereas account must be taken of all losses
related to the financial year or to a previous one. The origins of prudence, however, date much
further back.
Accounting regulators have preferred to write requirements that drive a conservative approach
to recording transactions, ie in simple terms, a ‘plan for the worst’ approach. The tendency has
been for losses to be recognized at the point at which management are aware of their probable
realization, whereas gains are recognized only when they are certain to be received. While some
see prudence as the opposite of imprudence – a clearly undesirable feature of any financial
reporting system – others see prudence as introducing bias into the financial statements.
There appears to have been a shift in attitudes during recent years. The idea that prudence
should be the dominant desirable attribute has passed and in its place the IASB have woven the
concept of neutrality. It is not entirely certain what this means though. Certainly, the deletion of
the term prudence from the chapter on qualitative characteristics was seen by some as a turning
point. At the same time, the previously fundamental qualitative characteristic ‘reliability’ was
replaced by ‘faithful representation’. The Conceptual Framework’s basis for conclusions states
that faithful representation ‘encompasses the main characteristics that the previous frameworks
included as aspects of reliability’. The section continues by stating that ‘substance over form,
prudence (conservatism) and verifiability, which were aspects of reliability in the previous
framework are not considered aspects of faithful representation’.
The IASB have been careful to design their Conceptual Framework to distinguish between the
following:

i the deliberate understatement of assets and profits, or overstatement of liabilities and


expenses; and
ii the adoption of a cautious approach in making the judgements necessitated by uncertainty so
that assets and income are not overstated and liabilities and expenses are not understated.

See supporting online content www.koganpage.com/accountingfm for a brief article which


draws attention to various practical drawbacks of (over-)prudence.

Expert view 2.8: Conditional versus unconditional conservatism


Academic literature also distinguishes conditional conservatism that results in asymmetric timeliness in the
recognition of good and bad news (the latter recognized earlier) and unconditional conservatism, which results
in systematic understatement of net assets. According to some academic literature, users find early recognition
of losses useful, as they are less frequently anticipated by the market than gains. There is a general agreement
on the usefulness of conditional conservatism, while unconditional conservatism is more contentious.

Disaggregation
To disaggregate means to separate into component parts. This principle is applied in accounting
whereby material assets and liabilities should normally be disclosed separately at their gross
amounts, rather than being netted off against each other. For example, netting off short-term
borrowings from a positive cash balance is not permitted.

Materiality
We have already discussed this concept, which might in its own way reinforce how fundamental
it is to financial accounting and accounting more generally. We introduced materiality as the
threshold quality. In other words, information is deemed to be material – by size or nature –
where its exclusion would impair an assessment of an entity’s position or performance. This rule
is advantageous in many ways, not least because it allows accountants to focus their attention on
the balances which are significant and, by default, to avoid absurd situations where immaterial
balances are being investigated at great cost by companies’ accountants.

WORKED EXERCISE 2.5 Mobius Inc (5)


Until this point, everything has been straightforward. You’re now ready to move to the next level, and the following
worked exercise shows how tangible non-current assets are accounted for. Though there are many standards, IAS 16
Property, Plant and Equipment is a good vehicle to allow us to discuss in more detail the prudence and accruals
conventions.
The previous exercise concluded at the end of week 2. Let us pick up the exercise at the start of week 3. The
closing position looked like this (ie opening statement of financial position as of first day of week 3):

Mobius Inc
Summarized Statement of Financial Position
As at the start of week 3
Assets
Non-current assets
Computer 500.00
Printer/scanner 100.00
Current assets
Cash at bank 1,450.00
Inventories
Raw materials 0.00
Finished goods 533.33
Trade receivables – amounts owed by customers 900.00
Liabilities
Loan from friend (500.00)
Trade payables – amounts owed to suppliers (800.00)
Net assets 2,183.33
Shareholder’s funds (capital and reserves)
Capital 1,000.00
Retained profits (weeks 1 & 2) 1,183.33
Shareholder’s funds 2,183.33

The following information is relevant to weeks 3 and 4:

Tangible non-current assets


(a) By the end of week 4, which marks the conclusion to your first full month’s trading, you have noticed signs of
wear and tear appearing on both of your tangible non-current assets – the printer/scanner and the computer. You
believe that the printer will continue effectively for 20 months, at which point it will be scrapped. The computer is
unlikely to be usable for business purposes after 40 months but you know a friend will buy it off you at that time
for $100.
(b) On the first day of week 3 you decide to buy a new motor vehicle which will be used exclusively for business
purposes. The useful economic life of the motor vehicle is estimated to be five years, at which stage the terminal
value would be $0. The invoice from the supplier showed the following costs:

$
Motor vehicle 19,500
Delivery charge 500
Additional extras:
Non-standard black matt paint job 1,000
Convertible roof function 2,000
Tank of petrol 150
Road tax (for the year) 200
Total 23,350

You take out a loan with a coupon rate of 10 per cent for the full amount from your bank to finance the purchase. The
interest is paid quarterly in arrears. The principal (ie capital amount borrowed) is due to be repaid in full in five years’
time.

Solution and explanation


The depreciation expense which appeared in Figure 2.1 now deserves separate attention.
Depreciation is an accounting estimate. The cash to acquire these assets has been paid on the day of acquisition.
We know that assets devalue over time and therefore it would be inappropriate (according to the prudence concept)
to hold them on the face of our statement of financial position at their purchase price until the day they are sold or
disposed of. Instead we spread the cost over the useful economic life of the asset (according to the accruals
convention). Depreciation is the measure of wearing out, consumption or other reduction in the useful economic life of
a non-current asset whether arising from use, effluxion of time or obsolescence through technical or market changes.

Printer/scanner
You have estimated that the economic life of the printer/scanner is 20 months. Therefore, at the end of the month,
you should show the asset as being reduced by one-twentieth of its value ie $100 / 20 years = $5 depreciation.
This depreciation charge gets netted off the carrying value (purchase price) of the non-current asset (ie $100 – $5
= $95) while the cost is taken to the income statement as an expense (reducing profit by $5). The amount which is
shown on the face of the statement of financial position is called the net book value (NBV).
This process would continue each month until the asset reaches the end of its useful economic life. In other words,
in each of the next 20 months you would charge $5 per month against the asset until all $100 had been consumed
(Table 2.3).

Table 2.3

Cost Depreciation Accumulated depreciation Net book value


Month 1 100 5 5 95
Month 2 100 5 10 90
Month 3 100 5 15 85
Month 4 100 5 20 80
Month 5 100 5 25 75
Month 6 100 5 30 70
Month 7 100 5 35 65
Month 8 100 5 40 60
Month 9 100 5 45 55
Month 10 100 5 50 50
Month 11 100 5 55 45
Month 12 100 5 60 40
Month 13 100 5 65 35
Month 14 100 5 70 30
Month 15 100 5 75 25
Month 16 100 5 80 20
Month 17 100 5 85 15
Month 18 100 5 90 10
Month 19 100 5 95 5
Month 20 100 5 100 0

The depreciation charge shown in Table 2.3 is taken to the income statement every month. The NBV (the final
column) appears on the face of the statement of financial position.

Computer
The treatment of the computer is similar. The only difference is that the asset has an estimated terminal (residual)
value.
You have estimated that the computer will last for 40 months, at which point it will be sold for $100. Therefore, now
we need only depreciate the asset down to its expected terminal value, as follows:

Therefore, at the end of the first month, the computer would be worth $490 ($500 – $10) and the depreciation charge
for this month (and every month until the end of the asset’s useful life) would be $10.

Acquisition of motor vehicle


IAS 16 Property Plant and Equipment states that the cost of an asset will include all costs in bringing the asset to its
required location and condition. As you can see, this is a subjective exercise. Is the convertible-roof function a
necessary improvement or adjustment to the asset to ensure it is suitable for the purpose?
Remember that the capitalized balance goes to the statement of financial position and will be written off over the
useful economic life of the asset. Any costs which you deem to be inappropriate to capitalize should be taken straight
to the income statement. We suggest the following treatment:

Capital Expense
$ $
Motor vehicle 19,500
Delivery charge 500
Additional extras:
Non-standard black matt paint job 1,000 1,000
Convertible roof function 2,000 2,000
Tank of petrol 150 150
Road tax (for the year) 200 200
Total 20,000 3,350

The carrying value of the asset would be recorded in the statement of financial position at a cost of $20,000 and
would be depreciated over five years. There is no terminal (residual) value. However, if depreciation is calculated on
a pro-rata basis, remember that you have owned this asset for only two weeks. Therefore:

Cost 20,000
Depreciation 20,000 / 5 years
= $4,000 per year
= $333.33 per month
= approx. $77 per week assuming 52 weeks a year (154)
Net book value 19,846

Bank loan to finance the purchase of the vehicle


The total loan required was $23,350. The loan is repayable in full in approximately five years’ time and therefore
would be classified as a non-current liability.
The interest payments (at 10 per cent per annum) are due quarterly in arrears. The annual interest charge would
be $2,335 ($23,350 * 10%). The quarterly charge would be $583.75. As we are accounting for only two weeks’ worth
of unpaid interest, the amount we need to accrue is $90 (calculated as: $2,335 / 52 weeks = (approx.) $45 per week).
As we owe the interest at the end of the first month of trading, it needs to be classified as a liability. Whereas the
loan balance is repayable in several years’ time, the interest is due within the next few months. Therefore this balance
should be classified as a current liability. A corresponding charge against profit needs to be made for the period to
ensure that we have matched the appropriate expense to the period.
The accounting entries would be as follows:

Increase the cost of tangible non-current assets in the statement of financial position by $20,000.
Charge $3,350 immediately to the income statement for the ‘additional extras’.
Increase the level of non-current liabilities by the same amount, ie $23,350. Remember, you needed a loan to
pay the motor vehicle supplier!
Charge $154 to the income statement related to the depreciation charge. The other side of the double entry
needs to be posted to the accumulated depreciation account. The net effect of this entry is to reduce profit by
$154 and reduce the carrying value of the motor vehicle by $154.
Finally, you need to accrue for the interest which is unpaid at the month end. $90 needs to be charged against
profit for the year and a current liability for unpaid interest needs to be shown in the statement of financial
position.

Extracts from statement of financial position and income statement for the
period ended week 4 (month 1)
NOTE: these are extracts, ie the statements are summarized and incomplete. We have simply highlighted the
balances that have changed as a result of the above transactions.

Mobius Inc.
Extract from the Statement of Financial Position
As at the end of week 4
Adjustment
Opening balance Closing
Note:
$ balance
$
Assets
Non-current assets
Computer 500 (10) 2 490
Printer / scanner 100 (5) 1 95
Motor vehicle ($20,000 [cost] – $154 [accumulated
19,846 3, 4 19,846
depreciation])
Current liabilities
Loan interest accrual (90) 5 (90)
Non-current liabilities
Bank loan (10%; repayable in 5 years’ time) (23,350) 3 (23,350)
Shareholder’s funds (capital and reserves)
Loss (3,609) 6 (3,609)
Mobius Inc.
Extract from the Income Statement
For the period ended DD/MM/YYYY (week 4)
Expenses
Depreciation – printer/scanner (5) 1 (5)
Depreciation – computer (10) 2 (10)
Motor vehicle costs (additional extras) (3,350) 3 (3,350)
Depreciation – motor vehicle (154) 4 (154)
Finance costs (90) 5 (90)
Profit/(loss) 0 (3,609) 6 (3,609)
Notes:

1. Depreciation charge on printer/scanner for the period of account.


2. Depreciation charge on computer for the period of account.
3. Motor vehicle ‘additional extras’ acquisition costs.
4. Depreciation charge on motor vehicle for period of account.
5. Interest paid on loan.
6. Transfer profit (loss) to statement of financial position (this transfer has been made purely for illustration
purposes and assumes that no other transactions took place during the period and we are closing off our
ledgers).

Three further property, plant and equipment issues


Let us take this opportunity to discuss briefly three further issues which you should be aware of
when accounting for tangible non-current assets:

recognition and subsequent measurement;


disposal of non-current assets;
alternative depreciation methods.

Recognition and subsequent re-measurement


We have simplified the example by assuming that all non-current assets are tangible and that the
assets are included at cost. The reality is that some non-current assets are intangible – thus
making the carrying value more difficult to quantify – and an entity has the choice whether to
revalue assets to their open market value (fair value) at the end of each accounting period. If
management choose to adopt a revaluation policy then this must be maintained and certain assets
should not be cherry-picked because of their value over other assets. In other words, if you have
some property that you believe has increased in value and other property that hasn’t, you cannot
choose to revalue only those assets which you believe it would be beneficial to revalue from a
financial position perspective.

Disposal of non-current assets


We have dealt with the acquisition of non-current assets but this is only a part of the story. These
assets can be disposed of or sold. If we take the example of the computer in the example above,
we are depreciating the asset down to a value of $100 over 40 months. That means that at the end
of the second (complete) year the asset would be held in the statement of financial position at an
NBV of $260 (ie $500 – $10*24 months). If you decide to sell the asset and find someone
willing to pay you $300, you would make a profit of $40. If you sell the asset for $100, you
would make a loss of $160.
At the point of disposal, the asset and its accumulated depreciation is completely written off
(leaving no trace of the asset in your statement of financial position) and the gain (loss) on the
difference between the sales proceeds and the NBV is credited (debited) to the income statement.
In a perfect world, there would be no difference between the sales proceeds and the NBV.
This would mean that your depreciation estimations were accurate. However, it is unusual not to
see gains or losses on disposal. These are simply a reflection of the level of under- or over-
depreciation over the ownership period. In other words, a gain means that you were overly
prudent in your estimation of the devaluation of the asset; a loss means that you were not prudent
enough. The final entry to reconcile proceeds and the NBV is a correcting entry.

Alternative depreciation methods


There are many commonly accepted methods for calculating depreciation. The reason that the
rules are flexible is because depreciation is an accounting estimate. Subjectivity is permitted but
the underlying ethos is that balances should be a true and fair reflection and that the costs of
ownership should be matched against the economic benefits which the asset provides.
Organizations should record their assets and liabilities, gains and losses as appropriately and
accurately as possible. Given that management are presumed to know more than other
stakeholders, it is only right that they should be responsible for choosing the depreciation policy.
In the case of depreciation we are deriving a measure to establish the level of wear and tear on
an asset, ie the consumption or other reduction in the useful economic life of a non-current asset
whether arising from use, effluxion of time or obsolescence through technical or market changes.
Therefore, it is easy to see why different methods of depreciation have evolved.
The two most commonly used methods to calculate depreciation are straight-line basis; and
reducing-balance basis. It is simplest to illustrate the differences between these methods through
a worked exercise.

WORKED EXERCISE 2.6 Depreciation methods


You purchase an asset for $500 and expect the residual value to be $0. In the first instance, you depreciate the asset
on a straight-line basis and in the second, you depreciate the asset on a reducing-balance basis:

1. The straight-line basis seeks to depreciate assets evenly over a period of time. You believe that the tangible
non-current asset has a five-year useful economic life. Therefore, each year you will charge $100 depreciation
against the asset ($500 / 5 years). Figure 2.3 shows the NBV of the asset over the time period.
2. The reducing-balance basis is commonly used for assets which lose value early in their economic lives and
less later. Examples of assets which you might choose to depreciate on a reducing-balance basis are motor
vehicles, high-tech goods and so forth.
To calculate the reducing-balance basis depreciation charge you need to know an appropriate percentage to
apply. Let us presume that the asset needs to be depreciated at 20 per cent on a reducing-balance basis (note:
the above example shows the asset being depreciated at 20 per cent on a straight-line basis). In the first year,
the depreciation charge is the same as under the previous depreciation basis because it is simply 20 per cent
of the purchase price. However, the second year, and all subsequent years, calculates depreciation based
upon the net book value (ie year 2 depreciation is $400 [NBV] * 20% = $80). The workings below and Figure
2.4 illustrate the difference.
Figure 2.3 An illustration of the impact of straight-line basis depreciation

NBV
$ $
Purchase price 500.00
Year 1 depreciation charge (20%) (100.00)
Net book value at end of year 1 400.00
Year 2 depreciation charge (20%) (80.00)
Net book value at end of year 2 320.00
Year 3 depreciation charge (20%) (64.00)
Net book value at end of year 3 256.00
Year 4 depreciation charge (20%) (51.20)
Net book value at end of year 4 204.80
Year 5 depreciation charge (20%) (40.96)
Net book value at end of year 5 163.84
Year 6 depreciation charge (20%) (32.77)
Net book value at end of year 6 131.07
… And so forth

Figure 2.4 An illustration of the impact of reducing-balance basis depreciation

Recording accounting information


As we draw towards the conclusion of this chapter, you might like to have a visual representation
of the way accounting books and records are maintained. We have simplified the record-keeping
process as far as possible mainly because the advent of computerized bookkeeping packages
means that many of these processes happen behind the scenes. At the press of a button these
systems allow one to print exception reports, a trial balance at a certain date, draft financial
statements and so forth. Systems also allow you to drill down to customer or supplier accounts,
locating single invoices if required; this would have been a long and exhaustive process not so
long ago! Of course, manual systems allow you to do the same; it is just that it is more time-
consuming. The flow chart in Figure 2.5 shows how information is collated and transferred
between ledgers, how it is subsequently summarized into a trial balance and then rephrased into a
set of financial statements.

Figure 2.5 Accounting books and records

Exercises: now attempt Exercise 2.5 on page 87

WORKED EXERCISE 2.7 Mobius (6)

Period-end adjustments
At the end of every accounting period, it is highly likely that a company will need to make a series of period-end
adjustments. The depreciation of non-current assets is an example of one of these. Frequently the adjustments are
correcting entries, accounting for estimates or adjusting the figures to an accruals basis. Let us once again consider
the statements of financial position and income for Mobius Inc as at the end of week 4.
The following information is relevant:

(i) Mobius Inc had a telephone line installed along with a broadband connection during the third week of trading. The
phone company invoices quarterly in arrears and you have not yet received your first invoice. This transaction
has not been recorded in your accounting records and therefore is not included within the financial statements
above. You have estimated that the usage during this first period of account will cost approximately $100.
(ii) On the last day of week 4, you found a suitable location to base the business. The monthly rental cost is $1,000
and the landlord required you to pay in advance. You paid in cash.
(iii) Mobius Inc sold $20,000 of product on credit terms. The total amount owed by customers (from all transactions)
at the end of the period was $4,900.
(iv) Mobius Inc bought a further $9,000 of raw materials on credit during weeks 3 and 4 of trading. Exactly $8,000
worth was converted into finished goods, of which three-quarters were sold. The remainder were held as
inventories at the end of the period. These were the only goods held at the end of the period as those finished
goods brought forward from weeks 1 and 2 were sold during week 3. Suppliers were owed $2,000 in total at the
end of the accounting period.

Solutions
(i) Mobius Inc reports on an accruals basis (not a cash basis). This convention asks that revenues, profits and the
associated costs incurred while earning them should be matched to the same period’s income statement.
Therefore, we need to charge the relevant proportion of the telephone cost against this period’s profit or loss. To
operationalize this accounting entry, we need a corresponding entry in the statement of financial position. In
other words, we need to charge the $100 telephone expense to the income statement and set up a current
liability for the same amount in the statement of financial position.
(ii) The $1,000 rental expense needs to be carried forward to the next period as this is the period in which it is
incurred. Let us break down this transaction into two parts. The first part is the cash transaction. When the cash
leaves the bank account, the corresponding entry would be to set up the rental expense in the income statement.
However, as we know, the expense should be carried forward to the next period of account and therefore we
have a charge of $1,000 against profit which should not be there. We need a period-end correcting entry as the
second part of the accounting transaction.
As the cash has been physically paid out, the ‘cash at bank’ balance cannot, and should not, be adjusted.
Instead, we need to include a current asset of a different type; in this case, a prepayment. By increasing assets
by $1,000 this leaves us with the other side of the entry to post to the income statement. All we need do is set
this off against the $1,000 charge which we had included as part of the cash transaction. This reduces the rental
cost in the income statement to $0 and sets off the prepayment gain against the cash loss, leaving the current
asset position at a net $0 position.

NOTE: The third part of the transaction is not dealt with here because it relates to the release of the prepayment. At
the end of the next accounting period, the rental cost will need to be charged against profits and the prepayment
removed from the statement of financial position (and possibly another set up in its place).

(iii) Revenue should be increased by $20,000. Trade receivables should be moved to $4,900. The cash balance
should be corrected appropriately to reflect the amounts received from these sales and the amounts invoiced but
not collected from prior periods.
(iv) The company has purchased $9,000 of raw materials, of which $8,000 was converted into finished goods stock
and three-quarters sold. Therefore, there are inventories left over at the end of the period. The closing balance of
raw materials should be shown as a current asset worth $1,000 and the closing finished goods inventories
should be included at $2,000.
Closing trade payables (current liabilities) should be shown in the statement of financial position at $2,000. We must
assume that all other balances have been paid.
The closing financial statements for the first four weeks of trading would be as follows:

Mobius Inc
Summarized Statement of Financial Position
As at the end of week 4 $ $
Assets
Non-current assets
Computer 490
Printer / scanner 95
Motor vehicle 19,846
20,431
Current assets
Cash at bank (1,450 [b/fwd] − 1,000 [2a] + 16,000 [3b] − 800 [4a] − 7,000 [4e]) 8,650
Inventories
Raw materials (1,000 [4d]) 1,000
Finished goods (533.33 [b/fwd] − 533.33 [4b] + 2,000 [4d]) 2,000
Trade receivables – amounts owed by customers (900 + 20,000 [3a] – 16,000 [3b]) 4,900
Prepayments (1,000 [2b]) 1,000
17,550
Current liabilities
Trade payables – amounts owed to suppliers (800 [b/fwd] − 800 [4a] + 9,000 [4c] − 7,000 [4e]) (2,000)
Loan interest accrual (90)
Accruals (100 [1]) (100)
(2,190)
Net current assets 15,360
Non-current liabilities
Loan from friend (500)
Bank loan (23,350)
Net assets (liabilities) 11,941
Shareholder’s funds (capital and reserves)
Capital 1,000
Retained earnings (weeks 1–4) 10,941
Shareholder’s funds 11,941

Mobius Inc
Summarized Income Statement
For the period ended DD/MM/YYYY (week 4) $ $
Revenue (2,850 [b/fwd] + 20,000 [3a]) 22,850
Cost of sales
Opening inventories 0
Purchases (666.67 [b/fwd] + 533.33 [4b] + 9,000 [4c]) 10,200
Closing inventories (1,000 [4d] + 2,000 [4d]) (3,000)
(7,200)
Gross profit 15,650
Expenses:
Utilities (200)
Printing, postage and stationery (50)
Web development costs (750)
Depreciation (5+10+154) (169)
Motor vehicle costs (3,350)
Telephone services (100 [1]) (100)
Rental costs (+1,000 [2a] – 1,000 [2b]) 0
Profit before interest and tax 11,031
Finance costs (90)
Profit 10,941

WORKED EXERCISE 2.8 Mobius (7)

Statement of cash flows


Mobius Inc is a simple business with no complex transactions and therefore converting the accruals-based financial
information to reveal the cash flows is quick and straightforward. Of course, being at the end of the first period of
trading also simplifies the preparation process. Note the format, in particular the categorization into three distinct
sections: operating activities; investing activities; and financing activities. This is designed to facilitate the
interpretation and analysis of the cash position.

Mobius Inc
Statement of Cash Flows
For the period ended DD/MM/YYYY (month 1) $
Cash generated from operating activities:
Cash generated from operations (note 1) 4,490
Interest paid (include either here or financing below) (90)
Dividends paid (include either here or under financing below) 0
Net cash flow from operating activities 4,400
Cash flows from investing activities:
Assets purchased (20,600)
Assets sold 0
Net cash flow from investing activities (20,600)
Cash flow from financing activities:
Issue of shares 1,000
Receipt of loan 23,850
Repayment of loan 0
Net cash flows from financing activities 24,850
Net cash inflow/(outflow) from activities 8,650
Opening net cash 0
Closing net cash 8,650
Note 1: Reconciliation of net cash from operating activities
Cash generated from operations
Profit for period 10,941
Interest paid 90
Depreciation 169
Increase in trade receivables (including prepayments) (5,900)
Increase in trade payables (including accruals) 2,190
Increase in inventories (3,000)
4,490

Exercises: now attempt Exercise 2.6 on page 88

Expert view 2.9: Statement of cash flows


The advantage of being asked to prepare a statement of cash flows (over the preparation of the other financial
statements) is that you are given the balancing figure – the closing balance of cash and cash equivalents.
Sometimes this might be a combination of balances (eg petty cash, bank overdraft, short-term cash and cash
equivalents) but ultimately, this makes the exercise easier to complete. Interestingly, you might like to note that
when completing this exercise in real life, computerized bookkeeping packages often struggle to produce an
accurate statement of cash flows unless the person responsible for inputting the data clearly labels items ‘cash’
and ‘not cash’.

COMPREHENSION QUESTIONS
1. Outline the content and purpose of the statement of financial position.
2. Outline the content and purpose of the statement of comprehensive income.
3. Outline the content and purpose of the statement of cash flows.
4. What is the accruals concept? Provide one example to show you understand this.
5. If an entity’s accounts are prepared on a break-up basis, what does this mean?
6. List and describe two methods of depreciation. For each, provide an example of an asset where it might be
appropriate to use that method.

Answers on page 89

Exercises
Answers on pages 90–95

Exercise 2.1: The components of financial statements


(a) The income statement shows:

Exercise 2.2 can be found on page 59

Exercise 2.3: Climb On!


Illustration 1.1 in Chapter 1 provided the following information:

For a long time, you have been wondering how to convert your passion for rock climbing into a business
opportunity. During a recent climbing trip you met Chris, a climbing-gear designer. He agreed to give you
100 chalk bags and 100 climb-oriented t-shirts for $4 and $5 respectively. You came to an agreement with
him that you’d make payments on an ad-hoc basis as the goods were sold. As soon as you arrived home
you listed the first 30 chalk bags and 50 t-shirts on an internet-based auction site. These sold within 3 days
of listing them for $8 and $10 respectively. Half of these customers paid immediately. You offered 20 day
terms on the sales and your past experiences tell you that customers tend to take full advantage of this
policy.

Required:
Based on this information, prepare a statement of financial position and an income statement.

Exercise 2.4 Goblin Combe plc


State which of the following items could appear as an asset on the statement of financial position of Goblin
Combe plc, a leading premium drinks business:

$150,000 of product sold during the year to Troillus Direct Inc under 40-day credit terms. The amount
remained outstanding at the end of the year and management believe that the amount will never be paid.
Goblin Combe plc holds $22 million of finished goods inventories as at the year-end. Of this amount,
$500,000 relates to a product which was banned from sale during the year. The directors have
ascertained that this particular product is highly effective as paint remover. A buyer has been found for
this and they are willing to pay $100,000 for the full quantity of this otherwise unsaleable stock.
A competing company produced a popular whisky called ‘Arbol’. Goblin Combe plc acquired the
company (and by default the ‘Arbol’ brand) at the start of the year for $30 million. The fair value of the
assets less liabilities, at that time, was estimated to be $10 million.
Goblin Combe plc hired a new corporate communication team. It is confidently expected that their
services will lead to an increase in profits by over $12 million per annum.
A product was developed by a rival company, Old Down Quarry Inc. The directors of Goblin Combe plc
decided to buy the exclusive rights to manufacture and distribute this product for the next four years at a
cost of $2 million. The new drink has already proved successful and sales have exceeded expectations.

Exercise 2.5 Climb On! (continued)


Trading continues apace for your new business Climb On! The products have proved to be popular and, seeing
this as your opportunity to seize the day, you decided to expand and grow the business.
Here follows a summary of your cash book, ie all cash transactions during the year to 31 December 2013:

Cash In Cash Out


Description $ Description $
Cash sales 32,000 Payments to suppliers 49,000
Receipts in respect of credit sales 106,000 Purchase of machinery 55,000
Capital invested (transfer from private bank account) 30,000 Rent and rates 7,800
Bank loan 45,000 Utilities 3,500
Interest received 200 Insurance 1,200
Sale of machinery 2,500 Telephone 300
Postage and packaging 200
Website development costs 1,500
Travel/climbing trips 10,500
General expenses 6,300
Wages (staff) 12,000
Drawings (your remuneration) 8,000
Interest paid 2,300
Balance carried forward 58,100
215,700 215,700

The following information is also available:

(a) The machinery was purchased on 1 April 2013. The estimated useful economic life of these assets is four
years. The residual value is estimated to be $nil. You may assume a full year’s depreciation in the year of
purchase but none in the year of sale.
(b) Some of the machinery quickly proved to be unnecessary and was sold during the year for $2,500. The
original cost was $5,000.
(c) Utilities bills of £400 were still owed as at 31 December 2013.
(d) Closing inventories as at 31 December 2013 were $14,000.
(e) Trade receivables as at 31 December 2013 were $10,500.
(f) Trade payables as at 31 December 2013 were $18,000.
(g) You need to provide for $1,500 of accounting fees as at 31 December 2013.

Exercise 2.6 Climb On! (continued)


Based upon your solution to Exercise 2.5, prepare a statement of cash flows for Climb On! for the period ended
31 December 2013.

Answers to comprehension questions


1. Statement of financial position (commonly referred to as the balance sheet), as the name suggests, is a
record of a company’s financial position as at a certain point in time. For this reason people think of it as
a financial photograph: a snapshot of the current worth of an organization.
2. Statement of comprehensive income shows how much profit the business has made during a specific
period of time. We told you that the ‘period of account’ is normally a year but large organizations are
often required to produce interim statements, ie bi-annually or quarterly.
3. Statement of cash flows shows whether a business has earned or spent cash (and cash equivalents)
during the year. The focus of the statement of cash flows is, as the name suggests, cash inflows versus
cash outflows. The statement is divided between three activities: operating; investing; and financing.
4. The accruals concept (often referred to as the matching concept or matching rule) states that revenues,
profits and the associated costs incurred while earning them should be matched to the same period’s
income statement. There are many everyday instances where a transaction would be recorded in
different accounting periods dependent upon whether you employed an accruals accounting system or a
cash accounting one. It is common for rent to be paid in advance, for example. In this instance, as at the
year-end you will have overpaid or prepaid the rental expense.
5. When an entity prepares accounts on a break-up basis it means that the management do not believe the
business will continue into the foreseeable future and that it has ceased to be a going concern. If this is
the case, all assets and liabilities are presented as though they were current.
6. There are several ways to account for depreciation of a non-current asset and it is the business’s
responsibility to choose that which is most fair and appropriate. The two most commonly used methods
are straight-line depreciation and reducing-balance basis. Straight-line depreciation reduces the value of
the asset in equal annual increments until the asset reaches its terminal residual value or is sold. The
reducing-balance basis uses the net book value of the asset and reduces this incrementally each year.
The former is used for assets such as fixed-term leases while the latter can be used for items such as
motor vehicles that tend to lose value early in their life cycle but hold value later.
Answers to exercises

Exercise 2.1 The component of financial statements

Exercise 2.2: Accounting equation adjustments


Exercise 2.3: Climb On!
Climb On!
Statement of Financial Position
Current assets
Cash at bank (Revenue * 50%) 370.00
Trade receivables (Revenue * 50%) 370.00
Inventories ([70*$4]+[50*$5]) 530.00
Liabilities
Trade payables ([100*$4]+[100*$5]) (900.00)
Net assets 370.00
Capital and reserves
Retained earnings 370.00
Climb On!
Income Statement
Revenue ([30*$8]+[50*$10]) 740.00
Cost of sales ([30*$4]+[50*$5]) (370.00)
Profit 370.00

Exercise 2.4: Goblin Combe plc


Definition reminder: an asset is a resource controlled by an entity as a result of past events and from which
economic benefits are expected to flow to the entity.

Under normal circumstances, the $150,000 would have been included as an asset and shown as part of
the trade receivables balance. However, it would appear that the amount is not recoverable (ie economic
benefits are not expected to flow to the entity) and therefore should be written off and not appear as an
asset at the year-end.
Accounting tends towards prudence and inventories should be included at the lower of cost and net
realizable value. Therefore, the final value of finished goods inventories in the statement of financial
position will be $21.6 million, ie $22m – $0.5m + $0.1m.
The assets acquired are worth $10 million but the cash being paid is $30 million. We know that cash will
be reduced by $30 million and assets will increase by $10 million. Therefore an accounting entry is
required for the difference between these two figures. Essentially the company is paying for the intangible
value of the entity, so taking the cost to the income statement as an expense would be inappropriate.
In accounting terminology this difference is referred to as goodwill. In accounting we tend towards
prudence and therefore there are rules which determine whether an intangible asset can be included on
the statement of financial position. In this instance, the difference between the consideration payable for
a business and the aggregate fair value of its identifiable assets less liabilities can be classified as
(purchased) goodwill and included on the statement of financial position as an intangible asset. At the
end of each accounting period an impairment test will highlight whether this intangible asset is being
included at an appropriate value.
Human capital is an example of an asset class which cannot be capitalized. Though there has been a
great deal of professional and academic debate about the ‘rights’ and ‘wrongs’ of this approach,
ultimately employees are not capitalized. Amounts paid to employees should be taken directly to the
income statement.
This is an example of an intangible asset which can be capitalized. The acquisition cost should be
matched against the expected life span. Annual impairment tests are required to ensure that that asset is
being carried forward each year at an appropriate value.

Exercise 2.5: Climb On! (continued)


Exercise 2.6: Climb On! (continued)
03
Financial analysis: Part I

OBJECTIVE
The objective of this chapter is to facilitate the development of financial information analysis and interpretation skills. The
emphasis of this first financial analysis chapter will be on undertaking appropriate forms of analysis, including calculating
key management ratios. This chapter will show how the results of the various techniques described can be employed to
help assess the position and performance of an entity.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Identify the main aspects of performance which can be evaluated through financial statements.
Apply horizontal, trend, vertical and ratio analysis to the financial statements contained within an organization’s
annual report.
Evaluate a company from the viewpoint of current and potential investors as well as other stakeholders.

KEY TOPICS COVERED


Aspects of financial analysis: profitability, liquidity, efficiency, solvency and investors’ returns.
The use of ratios and other techniques to analyse and appraise full sets of company accounts.
This chapter will contain mini case study exercises in which students must evaluate a company from varying
viewpoints.

MANAGEMENT ISSUES
Although it is useful that managers be able to compute ratios, it is more important that they are able to develop the
analytical skills through interpretation and analysis of financial statements.
Introduction
This chapter considers various forms of financial statement analysis, including:

horizontal analysis;
trend analysis;
vertical analysis;
ratio analysis.

We will combine these approaches in the chapter that follows. We will assume that undertaking
these forms of analysis is (relatively) new to you. When used appropriately, ratio analysis can be
a highly effective way of manipulating data into meaningful information. The key to undertaking
ratio analysis is to focus on the audience and their question(s). A current (or potential) investor,
for example, might want to know about the position, performance and financial strategy of the
organization. A lender might want to gauge the short-term liquidity – to gauge whether the firm
can make their obligatory interest payments – and the long-term solvency – to identify whether
there could be problems meeting repayment plans. On the other end of the continuum, an
environmental campaigner would probably have very different motivations for investigating the
annual report and financial statements. When calculating and interpreting key ratios they are
likely to be judging the firm from a social, environmental or ethical perspective. Therefore, the
ability to calculate a long series of financial ratios is good knowledge to have but a more useful
skill-set to develop is being able to identify the most appropriate ratios and then have the
capacity to interpret them.
The usefulness of financial ratios has long been established. Chen and Shimerda (1981) in
their paper ‘An empirical analysis of useful financial ratios’ have this to say:
Financial ratios have played an important part in evaluating the performance and financial condition of an entity. Over the
years, empirical studies have repeatedly demonstrated the usefulness of financial ratios. For example, financially-distressed
firms can be separated from the non-failed firms in the year before the declaration of bankruptcy at an accuracy rate of better
than 90% by examining financial ratios. In determining bond ratings, when financial ratios were the only variables used, the
resulting ratings were virtually identical with institutional ratings. There is one recurring question with the use of financial
ratios: which ratios, among the hundreds that can be computed easily from the available financial data, should be analysed to
obtain the information for the task at hand? (pp 51–60).

During this chapter we provide a list of financial ratios which we believe are the most useful for
you to know. Alongside each ratio we show how it can be calculated and a guide to help you
develop your analysis and interpretation skills. In an attempt to make this more real to you and to
bring the content alive, we have worked through the ratios of an airline.
We subdivide the chapter into five broad categories:

profitability;
liquidity;
efficiency;
solvency; and
investors’ returns.

Financial statement analysis for investment purposes


Investment-led financial statement analysis – buy, hold or sell – is predicated on the assumption
that investors are rational, risk averse and seek to maximize their returns in terms of the present
value of future cash flows. In other words, we presume that investors have an efficient frontier
where returns are traded off against risks, and investments will be rejected where the rewards are
not appropriate.
However, before an investment should be made in an entity, there are two fundamental
assumptions:

1. Buying shares on a stock exchange is not just an investment on paper, it is an investment in


a business.
2. Before you invest in a business, you should know the business.

Therefore, do not limit your assessment of position and performance to a numeric exercise. If
you are going to undertake an analysis exercise for a business in which you are either an investor
or considering making an investment, it is essential that you understand the industry and the
firm’s position in it. Among the issues which you should consider are:

the objectives of the organization;


the business strategy;
knowledge of the products sold/services offered;
knowledge of the products being developed;
your level of trust in the senior management team, and their vision;
the competitiveness of the industry and the position of the business within it;
the reputational capital of the business;
the political, legal, regulatory, social and ethical environment in which the business
operates;
the technology required to bring the product to market;
the reliance on skilled and unskilled employees and the competitiveness of the job markets;
and so on…

Other users and their needs


As stated earlier, investors are not the only user group who have financial information needs.
They are also not the only group who are interested in an entity’s financial statements. Table 3.1
sets out some broad motivations that these various user groups might have, alongside the type of
information that would satisfy their needs.
Table 3.1 User groups’ information needs

Potential motivations to
User group Example financial information requirements
consult the annual report
Concerned with the
stewardship of the entity. The
key issues are growth (historic What is the return on the investment (dividend and capital growth)
Investors
and potential), performance versus the level of risk taken?
(past, present and future),
position, risk and returns.
Employees would want to Is the company financially stable and will it continue into the foreseeable
Employees know whether the firm was future? Levels of borrowing and maturity would be useful information.
stable and solvent. How profitable is the business and the segment I work in?
Might be a motivation to Is the company profitable? If so, is there an issue of over-profitability?
investigate based upon How do prices compare to competitors?
Customers
fairness of prices, or maybe an Is the company (who is your supplier) solvent or should you be looking
over-reliance on one supplier. for an alternative?
Is the firm solvent and does it
Consider issues such as interest cover ratios, return on capital
Lenders have the short-term resources
employed, debt to equity ratios.
to meet interest payments?
Suppliers would be primarily
concerned with whether they
will be paid what they are
owed.
Suppliers They might also have Consider issues such as liquidity (short-term financial position), working
and other dependency issues, ie whether capital position and requirements, and solvency issues (medium- to
creditors the company who is their long-term financial position).
customer will continue to
operate into the foreseeable
future or whether they should
be looking for new customers.
For large entities, the annual report either includes or is accompanied by
a corporate social responsibility report. This is, of course, highly useful
to decision making about non-financial matters. The public, particularly
Is the firm operating in a
communities within which big businesses operate, will also be
manner that is
Public concerned with community matters, eg how do the company treat
environmentally, socially and
employees in terms of remuneration; will the company continue to exist
ethically responsible?
and invest in the local community; is it growing or contracting?
Therefore, standard measures of profitability and performance will also
be important to this user group.
The government and their
agencies will be concerned
with a range of issues such as
The questions being asked will depend upon the nature of the enquiry.
taxation, directors’
Government Levels of revenue are important to sales tax (value added tax [VAT])
remuneration, employment
and their concerns. Levels of operating profit will be relevant to corporation tax.
details, governance and so
agencies The solvency of the entity might be interesting when establishing the
forth. The annual report is
broader macroeconomic environment.
often the nominated repository
for such disclosure
requirements.

Horizontal analysis and trend analysis


Technique outline
Horizontal analysis invites you to make horizontal comparisons on a line-by-line basis. The
objective is to gauge relative levels of performance (or position) for a firm over a given period of
time. It is possible, of course, to undertake this analysis on raw numbers (horizontal analysis),
but more often than not, it is better to use the first period of account as your base position and set
to ‘100’ (trend analysis). You would then analyse future periods’ percentage growth against this
base position. Therefore, if revenue moved from $100 million in year 1 to $110 million in year 2,
revenue growth could be expressed as either $10 million (horizontal analysis) or 1.1 (or 110 per
cent of base; trend analysis) (ie $110m – $100m / $100m). Alternatively, if revenue fell during
year 2 to $75 million, this would be recorded as 0.75 (or 75 per cent of base), indicating a 25 per
cent decrease from the opening position.
Though carrying out this exercise can be time-consuming, it can also be rewarding in terms of
its ability to reveal patterns and trends, as well as irregularities and anomalies. It is not
uncommon for this analysis to be the first step for audit firms when they receive a new set of
annual financial information. It is a straightforward analytical review. Comparing figures on a
line-by-line basis can form the basis of discussions with the financial controller about the
performance for the period. This technique also has the capacity to immediately highlight any
exceptional movements in balances from previous years’ accounts, and thus those balances that
are higher risk (because of the possibility of misstatement). This can help to focus audit work on
areas of potential material error or misstatement.
When forecasting, this technique is often employed as a believability check and it is not
uncommon for people to talk about the hockey-curve effect, ie a graph depicting revenues
increasing in the shape of a hockey stick are likely to be ‘too good to be true’.
We have chosen to adopt Ryanair Holdings plc (hereafter, ‘Ryanair’) as a case study to
showcase these techniques. Over the past decade they have become a household name across
Europe and earned a reputation as a successful low-cost airline. By way of introduction, their
corporate website states the following:
Ryanair is the world’s favourite airline operating over 1,500 flights per day from 53 bases on 1,500 low fare routes across 28
countries, connecting over 168 destinations. Ryanair operates a fleet of over 290 new Boeing 737-800 aircraft with firm orders
for a further 13 new aircraft (before taking account of planned disposals), which will be delivered over the next year. Ryanair
has a team of more than 8,500 people and expects to carry over 80 million passengers in the current fiscal year.
(www.ryanair.com [April 2013])

Worked exercise 3.1 provides a decade’s worth of selected financial information – 2003 through
to 2012. Horizontal analysis reveals a strong growth profile.

WORKED EXERCISE 3.1 Horizontal analysis and trend analysis for Ryanair
Holdings plc

Table 3.2
Issues
The major issues with this kind of analysis are:

Presuming that balances are comparable between years:


– There are likely to have been changes in accounting policies, accounting requirements,
financial strategy and so forth that make direct like-for-like comparisons impossible.
Deciding which period to set as the baseline year:
– For example, setting 2009 as Ryanair’s baseline year would generate significantly less
useful results.

A note on financial analysis exercises


Remember that the simplest aspect of these exercises is completing the data collection exercise
and performing the mechanical calculations. The focus of any financial analysis should fall on
the ‘analysis’. Ensure that you do not simply ask yourself questions such as ‘how much’ or ‘how
little’ has balance X changed, but also ‘why’ and ‘so what’. Some preliminary questions you
might like to ask yourself are:

What does this information tell me about the company’s performance?


Has the company done well compared with other financial periods?
Has the company exceeded its own targets?
How does it compare with other companies in the same industry?
What are the future prospects of: i) the firm; ii) the local economy; iii) the global economy;
iv) customers and suppliers; v) political reform; vi) social, environmental and ethical
circumstances etc?

Vertical analysis

What is vertical analysis?


While horizontal analysis identifies trends across time periods setting a specific year as the base,
vertical analysis seeks patterns in the data on a year-by-year basis. In an income statement, the
revenue should be restated as ‘100’ and all subsequent figures are measured according to this
baseline. In the statement of financial position, use your (net) assets figure as the baseline. This
will mean, of course, you have two totals both adding up to ‘100’. This exercise will highlight
those balances which are significant (in terms of size) from those which are less so. Again,
auditors use this technique to identify material balances and areas of higher than average
potential risk.
Worked exercise 3.2 shows Ryanair’s statement of financial position as at 31 March 2012 and
income statement for the year ended 31 March 2012 plus one comparative period.

WORKED EXERCISE 3.2 Vertical analysis for Ryanair Holdings plc

Table 3.3

Income statements
For the year ended 31 March 2012 2011
$ % $ %
Revenue 4,390.2 100.0% 3,629.5 100.0%
Operating expenses:
Staff costs (415.0) −9.5% (376.1) −10.4%
Depreciation (309.2) −7.0% (277.7) −7.7%
Fuel and oil (1,593.6) −36.3% (1,227.0) −33.8%
Maintenance, materials and repairs (104.0) −2.4% (93.9) −2.6%
Aircraft rentals (90.7) −2.1% (97.2) −2.7%
Route charges (460.5) −10.5% (410.6) −11.3%
Airport and handling charges (554.0) −12.6% (491.8) −13.6%
Marketing, distribution and other (180.0) −4.1% (154.6) −4.3%
Icelandic volcanic ash related cost 0.0 0.0% (12.4) −0.3%
Operating profit 683.2 488.2
Other income/(expense):
Finance income 44.3 1.0% 27.2 0.7%
Finance expense (109.2) −2.5% (93.9) −2.6%
Foreign exchange gain/(loss) 4.3 0.1% (0.6) 0.0%
Gain on disposal of property, plant and equipment 10.4 0.2% 0.0 0.0%
Profit before tax 633.0 420.9
Tax expense on profit on ordinary activities (72.6) −1.7% (46.3) −1.3%
Profit for the year 560.4 374.6

Expert view 3.1: Reflections of other stakeholders


This is not a criticism of the firm, but while current and potential investors might be pleased to see profits rise,
how do you think employees will feel about these figures? Do you think environmentalists will be happy to see
increased numbers of passengers and an airline achieving higher levels of profitability?

Table 3.4

Statements of financial position


As at year ended 31 March 2012 2011
$ % $ %
Non-current assets
Property, plant and equipment 4,925.2 54.7% 4,933.7 57.4%
Intangible assets 46.8 0.5% 46.8 0.5%
Other non-current assets 153.0 1.7% 137.9 1.6%
Total non-current assets 5,125.0 56.9% 5,118.4 59.5%
Current assets
Inventories 2.8 0.0% 2.7 0.0%
Other assets 64.9 0.7% 99.4 1.2%
Current tax 9.3 0.1% 0.5 0.0%
Trade receivables 51.5 0.6% 50.6 0.6%
Derivative financial instruments 231.9 2.6% 383.8 4.5%
Restricted cash 35.1 0.4% 42.9 0.5%
Financial assets: cash > 3 months 772.2 8.6% 869.4 10.1%
Cash and cash equivalents 2,708.3 30.1% 2,028.3 23.6%
Total current assets 3,876.0 43.1% 3,477.6 40.5%
Total assets 9,001.0 100.0% 8,596.0 100.0%
Current liabilities
Trade payables 181.2 2.0% 150.8 1.8%
Accrued expenses and other liabilities 1,237.2 13.7% 1,224.3 14.2%
Current maturities of debt 368.4 4.1% 336.7 3.9%
Derivative financial instruments 28.2 0.3% 125.4 1.5%
Total current liabilities 1,815.0 20.2% 1,837.2 21.4%
Non-current liabilities
Provisions 103.2 1.1% 89.6 1.0%
Derivative financial instruments 53.6 0.6% 8.3 0.1%
Deferred tax 319.4 3.5% 267.7 3.1%
Other creditors 146.3 1.6% 126.6 1.5%
Non-current maturities of debt 3,256.8 36.2% 3,312.7 38.5%
Total non-current liabilities 3,879.3 43.1% 3,804.9 44.3%
Equity and reserves
Issued share capital 9.3 0.1% 9.5 0.1%
Share premium account 666.4 7.4% 659.3 7.7%
Capital redemption reserve 0.7 0.0% 0.5 0.0%
Retained earnings 2,400.1 26.7% 1,967.6 22.9%
Other reserves 230.2 2.6% 317.0 3.7%
Equity and reserves 3,306.7 36.7% 2,953.9 34.4%
9,001.0 100.0% 8,596.0 100.0%

Interpreting vertical analysis information


It should be straightforward to appreciate why this mode of analysis is both appealing and
informative. An analyst’s focus will be immediately drawn to the most significant balances. By
comparing with prior years’ vertical analysis, it can also be used as a form of horizontal analysis.
In the case of Ryanair, the key drivers of profitability are revenue, staff costs, depreciation, fuel
and oil, route charges, and airport and handling costs. Fuel and oil costs equate to more than one-
third of total revenue (2012: 36.5 per cent; 2011: 33.8 per cent) and are therefore of prime
importance to the end result. Thus, changes in the wholesale price of fuel – which has become an
increasingly volatile commodity – or a shift in the way the cost of fuel and oil is managed
(hedging programme) can have far-reaching (positive or negative) consequences for an airline
such as Ryanair.
The asset balances which stand out from the face of the statement of financial position are the
non-current assets (ie the aeroplanes) and the level of cash and cash equivalents. The horizontal
analysis reveals that Ryanair does not maintain an annual dividend payment policy and therefore
cash is accumulated or used to generate returns unless spent. The company requires a cash buffer
to protect itself during lean periods. The airline industry is particularly volatile and there are,
unfortunately, many instances over the past decade of entities which have not survived. The
dividend payment in 2009, however, probably highlights that the management believed that the
cash resources being held were far larger than were required and they had two options available:
either the cash needed to be reinvested in the business; or the shareholders needed to be
rewarded. The latter option was taken. The liability that is most prominent relates to long-term
debt which is required to purchase the non-current assets.
Please note that there are many other useful and insightful observations one could make based
upon the vertical analysis above. This is a method, however, which should be used in
conjunction with other techniques as it simply allows an analyst to undertake a quick appraisal of
key figures. The strength of the technique lies in its simplicity to perform and draw swift
conclusions. Its weaknesses stem from the same roots.

Ratio analysis

Introduction to ratio analysis


Ratio analysis, at least in theory, is a straightforward process. One simply needs to carefully
select a relevant numerator and divide it by an equally relevant denominator to produce a
calculation that has the potential to provide new insights. Once the result has been found, an
interpretation and further analysis should follow where applicable. Your calculation might
provide a ratio, a percentage, a number of times or a number of days, dependent upon the
relationship you have chosen to examine.
You should note at this stage the important consideration that attempting to interpret one ratio
in isolation is almost completely redundant. The purpose of the exercise is to draw an opinion
about relative levels of position and performance. The question we should be asking is, ‘relative
to whom or what’? In short, ratio analysis is most commonly used in order to:

evaluate current year performance and position;


compare performance and position across time periods;
assess whether target ratios have been met;
review firm-specific performance against the industry as a whole.

Expert view 3.2: Ratio analysis interpretation skills development


Practise, practise and practise! Read, read and read!
We are surrounded by financial information and news. Though we can set out ratios to learn and give you
some hints and advice on how to interpret the results, nobody is able to develop interpretation skills without
practice and without reading what others have written:

Ratios are tools, and their value is limited when used alone. The more tools used, the better the analysis.
For example, you can’t use the same golf club for every shot and expect to be a good golfer. The more you
practise with each club, however, the better able you will be to gauge which club to use on one shot. So too,
we need to be skilled with the financial tools we use. (Diane Morrison, CEO REC Inc)

As a word of caution and advice, we have found that it is common for students to learn a long list
of ratios without genuinely engaging with the material which they are being asked to examine.
This is perfectly understandable, especially for those who have not been required to perform this
type of analysis before. It is also a problem for those who are not actively engaging on a daily
basis with current affairs, particularly the business news. We strongly advocate that you pursue a
broader contextual understanding. This will facilitate a more rewarding experience and will
allow you to analyse the figures in a more precise and discerning manner.

Key ratios

Profitability ratios
Profitability ratios are intended to measure the performance of the entity. Less sophisticated
observers might focus exclusively on the bottom line, ie profit for the year and compare with the
previous year. Indeed, as Extract 3.1 shows, profitability is a headline grabber. Equally,
however, Extract 3.1 shows that the business reporter has dug deeper into the information to
understand why the levels of profitability have changed. Thus, though year-on-year profits are
important, relying on them as an assessment of annual (or ongoing) performance is somewhat
naive. Also, comparing actual profits between entities is not an especially useful measure given
the importance of relative size to levels of return. We present the three core profitability ratios
below alongside interpretation hints. These are:

return on capital employed;


gross profit margin; and
net profit margin.

Of course, we strongly advise that you drill down into the results and undertake further analysis
where appropriate. This can be done using other financial and non-financial performance-related
information as well as by employing other analysis techniques, eg horizontal analysis, trend
analysis or vertical analysis.

Extract 3.1: Profitability as a headline grabber

a) Profits rise
Google profits rise despite trend to cheaper mobile ads
Daily Telegraph (19 April 2013)
By Martin Strydom

Google profits rose 16 per cent to $3.35bn in the first three months of the year as revenue increased despite a
trend toward cheaper ads on smartphones and tablets.
Larry Page, the Google co-founder and chief executive, said: ‘We had a very strong start to 2013, with
$14bn in revenue, up 31 per cent year-on-year.’ The profits beat Wall Street’s expectation and Google shares,
which ended the official Nasdaq trading day slightly down, regained ground in after-hours trades to $777.
The number of paid clicks on ads posted at Google pages was 20pc greater than those seen in the first
quarter last year and up three per cent from the final quarter of 2012. Meanwhile, in a closely watched figure,
the cost per click for advertisers dropped four per cent, indicating a trend toward less expensive mobile ads.
Google executives highlighted the ‘big bets’ being taken by the company, ranging from Android mobile gadget
software to self-driving cars and Internet-linked glasses.

b) Profit warning
Mulberry pins third profit warning on thrifty tourists
Chief executive says Somerset-based luxury label needs to focus on tactical international advertising
The Guardian, Friday 22 March 2013
By Josephine Moulds and Simon Neville

Mulberry shares have plunged by nearly a fifth after the fashion and handbags business issued a third profit
warning in a year, blaming a fall in profits on lower spending by tourists in London after Christmas. The
Somerset-based luxury label, famous for its trademark Bayswater and Alexa bags – which start at £800 and go
up to £4,500 for versions in ostrich leather – said full-year profits would not meet market expectations due to
‘disappointing’ trading over the past 10 weeks. However, Mulberry’s chief executive, Bruno Guillon, who joined
a year ago, defended the performance and said the latest warning would not affect his international expansion
plans. He said: ‘We are still very confident with the strategy and will not be changing it. We need to focus on
increasing our international presence, particularly with tactical advertising, so that tourists coming to London
and Paris from China know about the brand.’
In October, bosses issued a ‘severe’ profit warning after lower than expected international retail sales and a
shortfall in wholesale revenue. Mulberry investors suffered losses of 22 per cent in June when the company
warned of a sales slowdown. Guillon said: ‘Sales in London have been particularly bad, with tourists not
spending in our stores.’ However, according to the Office for National Statistics most recent data, tourist
spending in the UK increased 11% in January to £1.24bn, although the number of foreign tourists fell 1 per
cent.
It is the latest in a series of disappointments for the brand, which has seen an incredible rise in recent years,
with shares hitting £23.90 in April last year. They closed at £10.26 on Friday. Mulberry is now targeting pre-tax
profits of £26m for the year to March, compared with expectations of £30.7m. Revenues are expected to come
in at £165m, against analyst forecasts of £175m. Wholesale revenues for the year are expected to be down by
about 15 per cent on last year, as a result of slimming down its operations and lower than expected ordering
during the season. It was the wholesale business that caused the October profit warning, after the company
focused on supplying department stores it felt were more in keeping with the brand’s premium position.
Despite increasing focus on the international market, Guillon said the business would keep its UK identity
and is opening a new factory in Bridgwater, Somerset, creating 300 jobs. However, the 20 stores opening this
year will all be overseas. Mulberry expects retail like-for-like sales to grow by 6 per cent this year. The company
insisted the order book for autumn/winter 2013 was building ‘satisfactorily’.

(a) Return on capital employed (ROCE)


It is common for investors to view the return on capital employed (ROCE) as the most important
measure of performance. Unfortunately, there are over 30 possible variations to the calculation of
this ratio, which can make comparisons problematic. We present the most commonly used
version, but when you are provided with this information, always maintain an enquiring mind.
There are also many closely related ratios, for example return on investment (ROI), return on
shareholders’ funds (ROSF) and accounting rate of return (ARR).
ROCE is calculated as:
What does this ratio reveal?
The ROCE weighs up the level of return generated during a period of account relative to the
amount of capital that has been provided. Sometimes the ROCE is used as a threshold measure
by management when deciding whether to accept or reject future proposed projects. It is also
commonly used when appraising historic performance. In other words, looking at this
relationship through the eyes of the capital provider, investors are attempting to gauge how well
their money has been spent and how well the resources have been managed. In extreme cases,
one could argue that an entity that returns less than the risk-free rate of return might be better off
liquidating the assets and investing the cash proceeds in 10-year US government treasury bonds.
It is not a truism to suggest that the higher the ROCE the better. This oversimplifies the
problem by ignoring context. A high ROCE does not necessarily mean that an entity is a good
investment. Remember that investors weigh up returns against risk. For example, an entity with
an average ROCE of, say, 10 per cent in the oil and gas extraction industry would be seen as a
relatively poor performer, whereas the same ROCE achieved by a property trust company, for
instance, would be seen as a healthy investment.
One could rightly suggest, however, that if the ROCE fell below the company’s investors’
expected weighted required rates of return, the investment would be deemed a poor one.
Corporate entities are looking to invest in projects that generate wealth. If the ROCE falls below
these required rates of return, the projects which are being undertaken are destroying wealth.
Occasionally the decision to take on lower-yielding projects is necessary, but to do so on a
regular and long-term basis increases the risk of capital extraction which would be detrimental to
the position of the firm.

Drilling down to interpret the changes in the ratio


A ratio provides a result about which one can make broad, descriptive and somewhat
meaningless statements. For example, as Worked exercise 3.3 shows, the ROCE has increased
from 7.2 per cent in 2011 to 9.5 per cent in 2012. An interpretation of this ratio should not be
limited to statements such as: ‘This change points towards a good year for Ryanair in terms of
returns to investors and their performance.’ Apart from anything, we don’t know whether this
does actually indicate a good year for the company. Before we could make this assessment we
would need to ask a number of follow-up questions, for example:

Were the results in line with targets?


How do the results compare to those from previous years?
Does this ROCE meet or beat market expectations?
How did competitors perform during the same period of account?
How has this percentage increase been achieved?
Is this ROCE sustainable?

If all you have to work with is the financial information itself, your analysis will be limited.
Nevertheless, you will be able to make some strides towards identifying trends, patterns,
strengths, weaknesses and potential areas of upside and downside risk. For accounting non-
specialists it can be difficult to see how the balances in the primary statements interlink, but it is
extremely helpful if you are able to engage with this idea. This is partly why we went through
the exercises in Chapters 1 and 2, to help you to see the interrelationship between assets and
liabilities, gains and losses.
There are two variables that determine an entity’s ROCE: profit; and capital employed. For
the ratio to move upwards like this (from 7.2 per cent to 9.5 per cent), one of the following must
have occurred:

Profit has increased and capital employed has decreased.


Profit has increased proportionately more than capital employed.
Capital employed has decreased proportionately more than profit.
Profit has stayed the same and capital employed decreased.
Capital employed has stayed the same and profit has increased.

In this example, profit has risen by almost 40 per cent while capital employed has risen by 6 per
cent. The weight in the formula is leant to the denominator (because of the volume of capital
employed). Nevertheless, these movements upwards are all significant. Further analysis shows
that a 21 per cent increase in revenue is matched to only an 18 per cent rise in costs. These
changes during the year have given rise to a sizeable movement in operating profit. The increase
in the amount of capital employed has arisen almost entirely because of the increase in retained
earnings, ie the amount of accumulated profit taken to retained earnings in the statement of
financial position as a result of a successful year’s trading. It is unsurprising that the chief
executive’s report on the period starts with the sentence: ‘Our results for the past year underline
the enduring strength of Ryanair’s ultra low fare airline model here in Europe’ (Ryanair annual
report, 2012, p 5).
If things had been different and capital employed had increased, we would need to drill down
into other balances. For example, it might be that non-current liabilities could have increased, in
which case you might want to see whether this was the result of borrowing to fund the
acquisition of new long-term assets. If it was used to buy aeroplanes or other revenue-generating
property, plant and equipment (PPE), we would expect to see an increase in non-current assets in
the statement of financial position as well. This would then impact on ratios such as asset
turnover. More importantly, we would expect to see increases in revenue as a result of asset
purchases, assuming the assets were put to use. Note, however, that there may be a lag given that
it is unlikely that these assets could be brought into immediate use. (It is not uncommon for
airlines to buy planes from other failed or failing airlines because they can get bargains through
financial distress gains but then ground them for a time until they find a suitable route.) Given
this information, you should now have a whole host of further questions which you would want
to investigate, for example:

How did the company manage to increase revenue by 20 per cent without seeing any rise in
non-current liabilities?
Were the planes previously underutilized?
Did customer demand drop off last year and pick up this year?
Were planes bought in previous years which have been brought into operational usage this
year?
Were planes acquired through non-debt-financed contracts, eg through operating leases?
Were there price rises this year?
And so forth…

Complications in the calculation

i) How should profit be defined?


When we refer to profit in the calculation of ROCE, it is usual to consider operating profit as the
most useful indicator of performance rather than the profit for the year or retained profits. This is
also referred to as profit before interest and tax (or ‘PBIT’). The rationale for using this figure
over others is that it provides a clearer message about current performance, and just as
importantly, current stewardship. Using this figure also aids greater comparability between
entities which are often subject to different financing requirements and taxation regimes.

Why exclude taxation costs?


Assuming that taxation regimes are fair (ie higher profits = higher tax charge), then relatively
high taxation costs should be interpreted as a reflection of better performance and, by extension,
all things being equal, good stewardship. Therefore judging performance based upon profits after
tax risks levelling out the playing field by inappropriately recognizing those companies with
lower tax bills based upon lower pre-tax profits. Of course, we acknowledge that tax is
controllable to a certain degree and sound financial planning alongside flexible accounting
policies and practices can reduce the annual tax-cost burden. If you choose to invest in a business
it would be unusual and unwise to invest in the management’s ability to manage tax costs rather
than their ability to generate operating profits.

Why exclude finance costs?


Interest expenditure normally dominates the finance costs category (although other costs are
included, eg gains and losses on derivative financial instruments held for trading). Loans are
normally matched to the term over which the assets are held. For example, an aeroplane is
thought to continue to generate revenues for approximately 23 years and loans are negotiated on
this basis. If the senior management team turnover rate is, say, one every five years or so, we
must ask whether it is fair and appropriate to judge new management against old loans.
There is also some debate over whether profit before interest and tax actually means profit
before interest payable and tax. Indeed, it is common for finance income to be added to the
operating profit number while finance costs are excluded from operating profit.

ii) What does capital employed mean?


Those responsible for providing capital to a corporate entity have the expectation that the
management will select and undertake value-generating activities which adequately and
appropriately reward the risks which they have chosen to bear. Capital provision, or project
financing, is considered to arise from one or both of these sources: equity and/or debt.

Equity can be provided through internal financing, such as accumulated reserves or


shortening the working capital cycle. However, it is far more common that the term equity
is used in reference to the raising of funds through the sale of share capital.
Debt is a term which is used to refer to all forms of borrowing – of which there are many.
Companies might, for example, apply for loans from banks or other lenders, take out
mortgages to help them buy land and buildings, long-term hire purchase agreements to
acquire plant, vehicles and machinery, sell loan stock to finance an expansion plan, and so
forth.
– Note that we have drawn your attention only to long-term debt financing, thereby ignoring
other forms of short-term debt such as a bank overdraft. In short, there are no rules for
inclusion or exclusion of short-term borrowings. We suggest that if a bank overdraft is used
as a long-term financing option, it should be included as debt under capital provided.

Capital provided can therefore be summarized as:

Equity + Debt

In relation to the statement of financial position, we suggest you use as a proxy for these two
categories:

Shareholders’ funds + Non-current liabilities

Occasionally you will see businesses average out the capital employed. We suggest that this is
practical when appraising single one-off investments (as the opening and closing capital required
will be easily to hand), but when your focus is on calculating the ROCE for the entity as a whole,
it is simpler to extract the closing (period-end) positions and use these.

Extract 3.2: Profitability explained


Dairy firm Graham’s sees sales and profits surge
Graham’s said the increased cost of raw milk, cream and fuel were ‘still proving challenging’
BBC online (28 October 2013)
New contract wins and significant brand investment have helped to lift sales and profits for Graham’s The
Family Dairy.
The Stirlingshire-based company report that sales climbed by 21 per cent to £68m in the year to the end of
March. Annual pre-tax profits more than doubled to just over £1m.
However, the company warned it still faced tough trading conditions, with continuing volatility in milk prices
and a squeeze on margins. Managing director Robert Graham said: ‘A focus on the Graham’s brand and
understanding the value of our brand has brought great success during this period, however as a business we
continue to face tough trading conditions.’ The increased costs of raw milk, cream and fuel are still proving
challenging. ‘New contract wins and improved internal cost management have led to strong sales growth.’
He added: ‘However our commitment to invest for the long-term in operations, plant and machinery in
challenging market conditions means that we are not currently seeing a true return on capital employed.’

Expansion drive
The company has been investing heavily in its brand name, including spending a six-figure sum on a TV
advertising campaign. In recent years it has expanded its product range to include butter, cream and ice cream.
The Bridge of Allan-based firm is planning further expansion after being granted a total of £630,000 earlier this
year through the Scottish government’s Food Processing, Marketing and Co-operation (FPMC) Award. The
cash will help the business to buy a site and install a yoghurt and cottage cheese line in Bridge of Allan and
build new lines for semi-skimmed milk processing and low fat spreads in Nairn.

Anglo American boss says profits ‘unacceptably poor’


Mark Cutifani is looking to shake up Anglo American
BBC online (26 July, 2013)
The boss of Anglo American, Mark Cutifani, has described his company’s profits as ‘unacceptably poor’.
The mining giant reported a 68 per cent drop in first-half profits, after it was hit by falling commodity prices,
as the global economic picture remained weak. Profits after tax fell to $403m (£262m) for the six months to the
end of June, compared with $1.254bn in 2012. Mr Cutifani promised to cut spending and halve the number of
pipeline projects. He also said the company would boost production at underperforming mines. ‘Our
performance at the operating level, compared to our budgets, has been unacceptably poor,’ Mr Cutifani said.
‘Over the last eight (quarters) only 11 per cent of operations are delivering consistently against their targets
– we have to up that,’ he added. He also said he wanted a ‘more disciplined approach to planning, to execution
and delivery’.
Mr Cutifani was appointed chief executive back in April, when he took over from Cynthia Carroll. However,
Anglo American did see some benefit from improvements in certain areas of production and a fall in the
currencies of the countries where it operates. As such, the company’s underlying operating profit only fell 15
per cent to $3.262bn during the first half.

Tough times
Anglo American is the smallest of the big international mining companies and its shares have not kept pace
with the likes of Rio Tinto and BHP Billiton in recent years. Anglo has had some big issues to deal with, most
recently labour unrest at its operations in South Africa and multi-billion-dollar cost overruns at its Minas Rio iron
ore project in Brazil. By slashing costs and pipeline projects, simplifying the company’s management structure
and improving the performance of its assets, Mr Cutifani is hoping to improve cashflow by $1.3bn within three
years.
He also intends to raise the return on capital employed (ROCE) to 15 per cent, from its current 11 per cent.
ROCE is a crucial figure mining companies use to measure the value of the assets they own.

(b) Gross profit margin


The gross profit margin is a key ratio which enables the analyst to assess trading performance
before deducting operating expenses. Or, in other words, the direct profit earned in relation to the
sales made. The ratio is calculated as:

It is rare that there are definitional problems like those shown in the calculation of the ROCE
above. The gross profit is calculated as:
We introduced cost of sales as being:

And, in many cases, this simple method is employed. However, the cost of sales is more
appropriately referred to as the cost of goods sold. The revised inference is that you are expected
to include the direct costs of production in this figure. The cost of goods sold includes not only
direct materials purchase costs, but direct labour costs and similar. Inclusion or exclusion of
costs ‘above the line’ (ie within cost of sales or above the gross profit line) between entities can
create difficulties when performing analysis. If you are undertaking sector (or industry) analysis
and you identify significant differences from one company to the next, we suggest you calculate
the net profit margin and ensure that this isn’t the result of an expense classification difference.
Most companies will disclose their gross profit figure. Airlines, however, cluster all of their
costs into operating expenses. This is unusual. It does mean that we are not able to calculate a
gross profit margin for Ryanair. Nevertheless, when analysing a change in this ratio, your first
challenge will be to work out why. There are only two variables:

Has revenue increased, decreased or stayed the same?


Has the cost of sales increased, decreased or stayed the same?

To help target your investigations, you might consider the following possible explanations as to
why the variation could have arisen:

Changes in the sales mix, eg:


– Were new products launched during the year?
– Were old products withdrawn from sale?
– Were high-margin/low-margin products more/less popular this year than in prior years?
Changes in sales price.
An increase/decrease in discounts offered to customers.
A change in inventories sourcing procedures or/and supplier(s).
Changes to the products and the associated raw materials requirements.
An increase/decrease in production process efficiency.
An increase/decrease in discounts being offered by suppliers.
Changes to cost classifications, from or to operating expenses.
An increase/decrease in inventory value write-offs due to obsolescence or other factors.
And so forth…

(c) Net profit margin


The net profit margin is a key ratio which enables the analyst to assess trading performance after
the deduction of operating expenses. It is a measure of operating profit in relation to revenue and
provides a guide as to how well the company has performed during the year. The ratio is
calculated as:

As discussed above (in relation to the ROCE), the calculation of the net profit margin can also be
adapted to take different measures of profitability. Most commonly, the net profit margin uses
operating profit (PBIT). This facilitates comparability and takes into account the costs which are
controllable by management.
Levels of profitability are inherently difficult to compare between entities and between years,
let alone between companies operating in different industries. Results will be dependent upon
managerial stewardship and firm performance, but will also relate to external events such as the
prevailing economic conditions and the degree of competition.
Ryanair’s net profit margin has increased from 13.5 per cent in 2011 to 15.6 per cent in 2012.
As with previous ratios, this observation requires further analysis and explanation. A naive
observer would be keen to make the point that this is an improvement on the prior year and
indeed, it might be. However, this statement is not verifiable without reference to other
information. You might like to consider the following questions:

What does your horizontal, trend and vertical analysis show you about the cost structure and
year-on-year line item comparisons?
Have costs risen in line with, above or below inflation?
To what extent is the movement in the profit margin a reflection of commodity price
movements which are less (or sometimes even un-)controllable, eg fuel costs?
How have other companies responded to the economic climate and have their costs moved
comparably to the entity under review?
Have there been any recognition or measurement accounting policy changes, eg changes to
depreciation rates?

Though each entity’s cost structure and strategy will be different, the level of change in certain
costs is generally worthy of separate enquiry when you are undertaking this form of analysis:

research and development;


depreciation and amortization;
pension costs;
employee costs, including staff training and development spend (if separately disclosed);
directors’ remuneration;
government grants.

WORKED EXERCISE 3.3 Ryanair’s profitability ratios


Extract 3.3: Ryanair Holdings plc: Chairman’s report, 2012
CHAIRMAN’S REPORT

Dear Shareholders,
I am very pleased to report a 25 per cent increase in profit after tax to a new record of €503 million. This was a
strong performance despite a €367 million rise in fuel costs which we managed to offset by a 16 per cent rise in
average fares. During the year Ryanair delivered a number of significant milestones:

We grew our traffic by 5 per cent to 76 million passengers.


We took delivery of 25 (net) new aircraft and we had a year-end fleet of 294 Boeing 737-800s.
We opened 6 new bases and 330 new routes bringing the total number of routes operated to over 1,500.
We improved our industry leading passenger service with better punctuality, fewer lost bags and less
cancellations.
We completed a share buyback of €125 million in fiscal 2012 and €68 million in April 2012, and the board
have proposed a dividend of €0.34 per share amounting to approximately €489 million subject to
shareholder approval at the annual general meeting. The combination of the second special dividend
(subject to shareholder approval) and previous share buybacks and dividends will mean that Ryanair has
returned an industry leading €1.53 billion to shareholders over the past 5 years.
Fuel costs as a proportion of our total operating costs have risen to 43 per cent in fiscal 2012. We are 90 per
cent hedged for fiscal 2013, at just over $100 per barrel and we are faced with a further €320 million increase in
our fuel bill, a total increase in 2 years of €687 million. Oil price rises and higher winter airport charges at
certain government owned airports will make it commercially sound to ground up to 80 aircraft rather than
suffer losses operating these aircraft during the winter when yields are significantly lower. Nevertheless, we still
expect passenger volumes in fiscal 2013 to grow by approximately 5 per cent to 79 million passengers.
In the airline industry, we yet again face another challenging year with significantly higher fuel prices and
with European government fiscal deficits resulting in austerity measures and leading to falling European
consumer confidence. As recessionary pressures continue we believe more carriers will exit the industry and
we intend to take advantage of those developments, as we have this year, when we opened a new base in
Budapest following the closure of Malev, and significantly expanded our operations at Barcelona and Madrid
following the closure of Spanair. We believe that the winners will be those airlines with strong balance sheets
(we currently have over €3.8 billion in cash), the lowest costs and a strong sustainable business model…
Notwithstanding the issues we face, the outstanding people at Ryanair continue to work hard on behalf of
shareholders to reduce our costs while at the same time delivering the lowest fares in Europe to our 79 million
passengers. As a result, we still expect to generate significant profits in fiscal 2013 although these are likely to
be lower than we enjoyed in fiscal 2012.

Yours sincerely,
David Bonderman
Chairman

Expert view 3.3: Financial analysis provided by the senior management


team
One of the benefits of modern corporate reporting is the provision of a large amount of self-analysis and
reflection from the senior management team. Extract 3.3 shows how Ryanair’s Chairman interprets the 2012
results. Of course, this text might be intended for self-promotion purposes and there are signs of impression
management. Nevertheless, what is written by these company representatives needs to be consistent with the
financial information which is reported elsewhere in the annual report. It is a fantastic learning tool for those
who are new to corporate reporting and financial analysis.

Liquidity ratios
Liquidity ratios are intended to provide users with an understanding of the short-term position of
an entity, where short-term means the next 12 months. A change in levels of short-term assets
and liabilities can also provide some assurances over solvency and the sustainability of current
levels of performance. Excessively high levels of current liabilities would be a red-flag issue for
investors. However, excessively high levels of current assets would be interpreted as a poor use
of resources. Achieving the correct working capital position balance (ratio of current assets
versus current liabilities) is as much an art as a science. Note that these measures can be
manipulated by window dressing, ie creating or holding assets, which you otherwise would not
have, at the end of an accounting period.
There are many liquidity ratios which could be produced; however two key ratios which you
should start with are the current ratio and the acid test ratio (otherwise known as the quick ratio).
(a) Current ratio
The current ratio is calculated as follows:

The result is expressed as a ratio, but can also be seen as a level of coverage. You might see
references to target ratios. More specifically, in the past a current ratio of 2 : 1 and a quick ratio
of 1 : 1 were seen as being ideal. The presumption is that an entity can remain liquid if it has the
ability to cover its immediate bills two times over if called upon to do so. Times, however, have
changed and the perceptions of over/under-resourcing have also shifted. Management are
frequently appraised on their ability to extract the maximum value out of their short-term liquid
resources.
Commentators who hold to the belief that there is an optimum ratio of current assets to
current liabilities often neglect to mention the pay-off between liquidity and profitability. If one
has finite inventory production capabilities and makes the conscious decision during a period of
account to stockpile inventories, then (assuming all other things are equal) the following would
occur:

current assets will be higher at the end of the period;


thus, the current and quick ratios will increase;
but the amount of inventory sold will be lower;
and therefore profits will be lower.

As Worked exercise 3.4 shows, the proportion of current assets held versus current liabilities in
2012 by Ryanair was 2.14; or, current liabilities were covered by current assets 2.14 times. It is
common for an increase in the ratio to be interpreted as an indicator of a better position. As with
the profitability ratios, however, liquidity ratios cannot be appropriately or accurately
commented upon without taking into account the drivers of the change and the broader context in
which the change has occurred.
In the case of Ryanair, the increase has been driven mainly by a net increase in cash and cash
equivalents. The fair value of short-term derivatives holdings has dropped (assets and liabilities).
Though this could potentially have a material impact on fuel costs and other hedged expenses,
the impact on the liquidity ratios is not significant. The company was holding over €3 billion in
cash at the end of 2012. Generating this amount of cash must be viewed as positive. However,
not being able to spend it on operations might be seen as a weakness hence why the company has
returned approximately $1.53 billion to shareholders over the past five years and the promise that
further investments in aircraft are likely to follow in 2014/15. Also, in a broader context, one
should not forget that this increase in cash and rise in profitability has occurred during a period
of economic uncertainty, turmoil and recession. Therefore, an important point which we are
neglecting to mention is the relative performance of the firm during this period in comparison to
other airlines, and more specifically, other low-cost airlines.
(b) Acid test (quick) ratio
The acid test ratio is similar to the current ratio. The denominator remains the same (ie current
liabilities) but the numerator, current assets, is adjusted for inventories. The acid test ratio is
calculated as follows:

Current assets minus inventories


Current liabilities

The rationale for removing inventories from current assets is that they are the least liquid of
assets within the class. When businesses are faced with an immediate need for cash and face
genuine bankruptcy risk, their inventories are worth very little. When an entity sells inventories
as a going concern, they are almost always worth more than their cost. When an entity is being
wound up, the inventories held are normally worth significantly less than cost as buyers perceive
the opportunity to buy at a bargain price, plus they are aware that continued spare-parts
production will likely cease and maintenance will not be as readily available if the company does
not survive. This phenomenon has been shown to affect companies across a wide range of
industries and size does not mean you get treated differently, for example MG Rover (motor
vehicles) and Lehman Brothers (financial services).
Inventories held for airlines are typically low, especially airlines like Ryanair which specialize
in low-cost short-haul journeys. Ryanair carry a small amount of on-board food and duty-free
goods which are intended to be sold in-flight. Therefore the impact on their liquidity position is
immaterial. It might be interesting to note that when airlines are faced with bankruptcy concerns,
it is more common for competitors or new entrants to attempt to take advantage of their position
by buying the struggling firm’s non-current assets – planes and landing slots – at heavily
discounted prices (rather than their inventories).

WORKED EXERCISE 3.4 Ryanair’s liquidity ratios


Current ratio:

2012 2011
€m €m
Current assets 3,876.0 3,477.6
Current liabilities 1,815.0 1,837.2
3,876.0 3,477.6
1,815.0 1,837.2
=2.14 : 1 =1.89 : 1
(or 2.14 times) (or 1.89 times)
Current ratio:

2012 2011
€m €m
Current assets 3,876.0 3,477.6
Less − −
Inventories 2.8 2.7
Current liabilities 1,815.0 1,837.2
3,873.2 3,474.9
1,815.0 1,837.2
=2.13 : 1 =1.89 : 1
(or 2.13 times) (or 1.89 times)

Efficiency ratios
Efficiency ratios are calculated to provide analysts with information about the utilization of
(short-term) resources. This broad definition means that there are many ratios which could be
calculated and many relationships which might be worthy of detailed attention. We would urge
you to start with three basic working capital ratios and investigate further if required. These are
as follows:

trade receivables collection period;


inventories holding period;
trade payables payment period.

The aim of the exercise is to gauge whether the entity under consideration has managed their
resources more or less efficiently than i) in previous years and ii) their competitors As with other
ratio categories, however, the interpretation of the results will depend upon context, eg:

the industry the firm operates in;


the working capital management strategy (aggressive or defensive) adopted by the entity;
the state of the economy;
competitors’ approaches to working capital management;
credit rating and availability of short-term credit;
and so forth…

The application of the prudence concept in financial accounting means that results remain
broadly comparable year on year. The inventories are valued at the lower of cost and net
realizable value, the trade receivables figure is shown net of balances owed but which are not
likely to be received, while trade payables includes amounts owed to suppliers which are
obligated liabilities of the entity.
Other ratios you might consider that could be used to explore efficiency include:

inventory turnover ratio (Cost of Sales / (Closing) Inventories);


average inventories turnover ratio (Cost of Sales / Average Inventories [opening inventories
plus closing inventories divided by 2]);
non-current assets turnover ratio (Revenue / Non-current assets);
revenue per employee (or any other non-financial cost driver);
revenue per aeroplane (or any other major asset class).

Extract 3.4: A profitable company with poor liquidity – what’s the worst
that can happen?
Wet ‘N’ Wild waterpark enters administration
BBC (online) 17 October, 2013
Wet ‘N’ Wild was opened in North Shields in 1993
A waterpark in North Tyneside has gone into administration, putting nearly 70 jobs at risk.
Wet ‘N’ Wild in North Shields went into administration on Monday despite being profitable ‘for much of the
year’.
The indoor waterpark opened in 1993 and employs 69 people, the majority of whom will be made redundant,
administrators PwC said.
A spokesman for PwC claimed it had ‘no alternative’ but to close the park and was seeking interested
buyers.
The company has been working with North Tyneside Council to pay off tens of thousands of pounds worth of
business rates and other liabilities, the BBC understands.
The council has not been involved in the administration process.
Toby Underwood, joint administrator and partner at PwC, said: ‘Despite operating profitably for much of the
year the business faced liquidity issues over the forthcoming months.
‘Unfortunately, with the quieter winter trading period upon us we have had no alternative but to close the
waterpark with immediate effect and make the majority of employees redundant.
‘We will continue to maintain the waterpark while seeking a sale.’

(a) Trade receivables collection period


The trade receivables collection period is calculated as follows:

A note on the calculation of trade receivables collection period


Note that the credit sales figure is rarely disclosed in an annual report. The headline figure,
revenue, however, is almost always shown separately and can be used as a proxy for credit sales.
Sometimes this is disaggregated into categories, which might be helpful for your analysis (for
example, BMW Group plc divide revenue between motor vehicle sales and financial services
revenue [interest received on loans offered to purchase BMWs by BMW Group plc]). You
should be aware that substituting revenue for credit sales might distort your analysis. Though the
majority of companies trade exclusively through short-term credit agreements (eg invoices
demanding payment is received within 28 days), this is not always the case. If you are
investigating the ratios of a supermarket, for example, the majority of customers pay in cash at
the checkout/till. These businesses permit only a handful of customers to buy ‘on account’. If, in
this case, you use revenue to calculate the trade receivables collection period, the result produced
will drastically shorten the real collection period.
Some argue that it is more appropriate to use average trade receivables than closing trade
receivables. They suggest that this tells you how long the average customer takes to pay. Both
closing and averaging have weaknesses. If it is obvious that they produce significantly different
results, a reconciliation alongside an explanation would be required.

Interpreting the trade receivables collection period


Though there are benefits to both buyer and seller, analysts tend to view trade receivables as
short-term interest-free loans to customers. The longer you allow your customers to pay, the
longer your working capital cycle. In turn, the greater the period of time is that you must spend
without cash. In this regard, there is an opportunity cost attached to trade receivables, ie an entity
could be doing something more valuable with their cash, for example investing it profitably in
operations.
In short, the trade receivables collection period is designed to indicate the length of time (in
days) it takes for credit customers to pay for the products or services they purchased. Commonly
a shorter collection period compared to prior years and industry averages is better. This means
that you are recycling your cash quicker. Assuming that the entity is profitable, this also means
that you are contributing to the generation of cash more effectively. In other words, you are using
your resources more efficiently. A shorter(-ening) trade receivables collection period could
indicate improved credit collection management, eg careful selection of creditworthy customers,
chasing late payments more vigorously, and so forth.
There are, however, several problems with this naive interpretation:

The nature of the calculation means that weight is added to larger customers. Thus, if one
large customer pays faster (slower), it will skew the result.
If several smaller customers are not paying, it would be difficult to pick up from this simple
ratio calculation; and yet this might be information important to your analysis and to the
business more generally.
One might be suspicious of the policies and processes of a company with a shorter than
average trade receivables collection period.
– A business that maintains a demanding (pushy?!) cash collection department/strategy runs
the risk of alienating customers, especially if the operational area is competitive and
customers could go elsewhere for similar products or services.
– Offering large discounts might speed up collection but there would be an offsetting impact
on profitability.
A shorter than average number of collection days might be due to an underlying weakness
in the calculation, eg the inclusion of cash sales in the revenue figure.

Unfortunately, Ryanair’s trade receivables collection period provides an almost meaningless


result (see Worked exercise 3.5). On the face of it, the calculation shows that the number of
outstanding days is low (around four days). This would ordinarily be interpreted as an extremely
efficient use of short-term resources. When this is analysed in context, however, this conclusion
lacks robustness. The majority of Ryanair’s customers are not offered a credit settlement period.
Instead, customers pay when they book their flights, ie by cash in advance. As we do not have
the disaggregated revenue figure, it is impossible to say how long customers who are allowed
credit take to settle their obligations. Providing some reassurance that the risk of default should
not be a material concern for investors, the company does separately disclose that ‘no individual
customer accounted for more than 10% of [their] accounts receivable at March 31, 2012, March
31, 2011 or at March 31, 2010’ (Ryanair Holdings plc, Annual Report 2012, p 155) and only 0.2
per cent of this balance was deemed to be impaired, or irrecoverable.

(b) Inventories holding period


The inventories holding period is calculated as follows:

Inventories are frequently a significant asset class for businesses and represent a large amount of
short-term resource being tied up. The inventories holding period tells the analyst how many
days the business is keeping their average item of stock. Given that any holding of inventories
represents tied-up cash – and by extension the lost opportunity to use that cash for value-creating
activities – then the shorter the inventories holding period, the better.
Interpreting yearly differences should be straightforward. The following might explain
movements:

a variation in sales mix;


a change in products or services offered;
commodity price movements;
supply chain changes;
a growth (decline) in demand;
improved (worsened) inventories control management;
high (low) levels of obsolete stock;
a change in buying strategy, eg bulk purchasing (to take advantage of discounts), or more
cautious buying policies;
a period of firm-specific growth;
and so forth.

There are two issues that prevent comparability between entities:

1. There are different inventory demands between industries and this will lead to large
differences in the result of the inventories holding period calculation. For example,
supermarkets stock short-life products by and large and therefore their inventory days will
be relatively low, whereas house-building companies might have long inventory days given
the nature of their business.
2. Not all business inventory management strategies are alike, either within or between
industries. Some companies will adopt sophisticated inventory management practices while
others will lag behind. Every entity will weigh up the costs versus the benefits of
introducing advanced inventory control systems innovations and make their choice
accordingly. For example, just-in-time (JIT) systems are popular in the motor vehicle
production industry. This involves having high-quality raw materials delivered by the
supplier, as the name suggests, just in time for production! Note that this system is not for
everybody, though, as there are many potential problems that could arise, for example a
breakdown in the relationship with supplier(s), unexpected production discrepancies,
quality issues, late changes to orders, priority orders and so forth.

If we look more closely at Ryanair’s inventory holding period, the first thing to note is the oddity
in the calculation. Most companies disclose their cost of sales, but airlines classify all expenses
as operating expenses. We have therefore been forced to find a substitute. There is a strong
temptation to use only maintenance, materials and repairs, but a more balanced view might be to
use operating expenses as a whole. Both calculations are shown below. The differences are
sizeable but whichever way we look at them, inventories holding is not a material risk for
Ryanair. The numerator, ie the inventories balance, represents less than a third of 1 per cent of
the entity’s total assets and less than half of 1 per cent of profit after tax.
This was not always the case, however. Inventories held by the company were reduced
significantly during 2005. The inventory holding period at that time moved from over 12 days in
2004 to below one day in 2005. The company now only holds essentials in stock, for example a
limited amount of food available for on-board snacks and a small selection of duty-free products
which are intended to be sold mid-flight. As Extract 3.5 shows, on-board food is not a trivial
matter for airlines. Bean counting is an oft-used derogatory term for accountants, but this article
brings the expression to life as well as the importance of the function to a successful business.
The report highlights one airline, Delta, who saved $210,000 per annum simply by cutting one
strawberry from their salads.

Extract 3.5: Airlines on-board food


Beyond Mile-High Grub: Can Airline Food Be Tasty?

Published March 10, 2012 in The New York Times


By Jad Mouawad

Last year, Delta hired Michael Chiarello, a celebrity chef from Napa Valley, to come up with new menus for
business-class passengers flying on transcontinental routes – New York to Los Angeles and New York to San
Francisco. It was not the first time that Delta had worked with a renowned chef. The airline has served meals
created by Michelle Bernstein, a Miami chef, since 2006 in its international business class.
‘Our chefs are like portrait painters,’ Mr Wilander says. ‘They can get pretty creative. But we need to
translate that into painting by numbers.’ That process began last May, when Mr Chiarello met with executives
and catering chefs from Delta at a boxy industrial kitchen on the edge of the San Francisco airport to
demonstrate some of his recipes. Among the dozens of dishes he tried were an artichoke and white-bean
spread, short ribs with polenta, and a small lasagna of eggplant and goat cheese.
‘I am known for making good food, and airlines generally are not,’ says Mr Chiarello, who is also the author
of a half-dozen cookbooks, the host for a show on the Food Network, and a former contestant on ‘Top Chef
Masters’ and ‘The Next Iron Chef.’ ‘I probably have a lot more to lose than to gain doing this.’
Huddled around him, white-toqued chefs from Delta and its catering partners weighed each ingredient on a
small electronic scale, took scrupulous notes and pictures and tried to calculate how much it would cost to
recreate each dish a thousand times a day.
It took Mr Chiarello six months to come up with the menu. He tested recipes, picked seasonal ingredients,
considered textures and colors and looked at ways to present his meals on a small airline tray. Then Delta’s
corporate chefs had to learn his way of cooking and serving. Bean counters – the financial kind – priced each
item. Executives and frequent fliers were drafted to taste his creations.
There were a lot of questions. How should cherry tomatoes be sliced? (The answer: Leave them whole.)
What side should a chicken fillet be grilled on? (Skin first.) How many slices of prosciutto can be used as
appetizers? (Two large ones, rather than three, struck the balance between taste and price.)
For airlines like Delta, these are not trivial matters. A decision a few years ago to shave one ounce from its
steaks, for example, saved the airline $250,000 a year. And every step of kitchen labor increases costs when
so many meals are prepared daily. An entrée accounts for about 60 per cent of a meal’s cost, according to
Delta, while appetizers account for 17 per cent, salads 10 per cent and desserts 7 per cent.
Delta also calculated that by removing a single strawberry from salads served in first class on domestic
routes, it would save $210,000 a year. The company hands out 61 million bags of peanuts every year, and
about the same number of pretzels. A one-cent increase in peanut prices increases Delta’s costs by $610,000
a year.
Others are catching on. United Airlines said in February that it would upgrade its service to first- and
business-class passengers and would change the way it prepares meals ‘to improve the quality and taste.’ It
also said it would start offering a new ice cream sundae option with a choice of six toppings on international
flights. On domestic flights, premium passengers will get new snacks, including warm cookies.

(c) Trade payables payment period


The trade payables payment period is calculated as follows:

A note on the calculation of trade payables payment period


The trade receivables collection period, as discussed above, demands that revenue is divided by
credit sales and multiplied by 365 days. However, the credit sales figure is rarely provided and
therefore we need to find a substitute. Though revenue is not a perfect replacement, it is normally
a satisfactory proxy. The same problem exists for credit purchases. This figure is rarely publicly
disclosed. Therefore we need to find a substitute. Commonly, one would use the cost of sales
figure. This is because in its simplified state, the cost of sales is: (i) the product of opening
inventories plus purchases minus closing inventories; and (ii) it is unusual for a large business
not to make purchases on credit terms.
The complications for analysing Ryanair’s efficiency ratios (see Worked exercise 3.5) are
that: firstly, expenses are classified as operating expenses; and secondly, trade payables could
relate to any of these operating expenses presented. Therefore, we have used operating expenses
as a substitute for cost of sales. Note that this is unusual.

Interpreting the trade payables payment period results


If one views holdings of current assets, such as trade receivables and inventories, as reducing an
entity’s cash resource, then current liabilities such as trade payables can be seen as an offset to
these effects. While trade receivables are essentially an entity making interest-free loans to
customers, trade payables are the reverse, ie suppliers granting interest-free loans to the business.
Therefore, it is common to believe that a higher trade payables payment period is preferable.
However, when the number of days a company takes to pay their suppliers increases beyond a
certain level, significant additional risks could arise, mainly resulting from behavioural issues
and the loss of supplier goodwill. A relationship breakdown could bring serious problems; for
example, a supplier who is made to wait disproportionately long before payment could penalize
the slow-paying entity by providing poor-quality product. In a worst-case scenario, they might
simply stop supplying products or services at all and poison the market against the slow payer.
In the case of Ryanair (see Worked exercise 3.5), the trade payables payment period has
remained fairly constant at around 17 to 18 days. Note that in 2003 this was as high as 41 days.
As with the other working capital ratios, it is difficult to interpret what this actually means. Also,
though the balance is material (€181 million: 2012), it is not at the level which would cause an
investor grave concern. The movement of €30 million seems like a large shift but in the grand
scheme of the business, this is insignificant.

WORKED EXERCISE 3.5 Ryanair’s efficiency ratios


Solvency ratios
We explained that liquidity ratios are intended to provide users with an understanding of the
short-term position of an entity. Solvency ratios, on the other hand, allow users to evaluate how
the entity is positioned for the medium and long term. Without some form of financing, it would
be impossible for a company to invest in projects which generate value. Thus, their going
concern status would be impaired. The investment cycle can be envisaged as shown in Figure
3.1.

Figure 3.1 Financing a new project

Expert view 3.4: Debt – the company’s perspective versus the


shareholders’ perspective
From a company’s perspective: Debt is a preferable financing option because, by and large, it is cheaper to
service than equity and because of the tax relief allowed on interest payments but not allowed on dividend
payments (if you wish to prove this to yourself, just think about the order of the income statement and note that
the tax charge occurs after the interest charge).
From a shareholder’s perspective: Debt financing increases the (perceived) risk as it normally has all, or
some, of the following characteristics:

Interest payments are an obligation. (Dividends are discretionary.)


Debt is secured. For example, non-repayment of mortgage payments on a building can lead to the
asset(s) being repossessed by the lender and sold to settle the liability. (Shares are rarely secured over
assets.)
Debt has a finite life, ie there is a redemption date. (Equity is a permanent investment.)
Equity providers rank last in a winding up.

The primary solvency ratios consider relative levels of debt and equity as this ratio sheds light on
the way the business is financed. It is argued that as an entity’s levels of debt rise, equity
providers (shareholders) require a greater return to offset this financing risk. The graph shown in
Figure 3.2 highlights the relationship between the cost of debt and the cost of equity as financing
risk (ie proportionately higher borrowings) is introduced into the business. The cost of debt is the
weighted average required return of providers of debt finance. The cost of equity is the required
return of investors with an equity stake in the business. Note that at extreme levels of financing
risk, the required returns of both debt providers and equity providers increase drastically. This
reaction builds in the compensation required for the exposure to bankruptcy risk.

Figure 3.2 The costs of financing a new project


Despite this knowledge, it is always difficult for an analyst to compare and interpret a
company’s level of financing. Generally speaking, we would state that a careful balance between
debt and equity needs to be achieved. A high gearing ratio (the proportion of debt to equity)
would suggest a greater exposure to solvency risk. However, the critical question is: ‘what is
high?’ Also, a low gearing ratio would suggest that the comparative cost advantage that debt
offers has not been appropriately taken advantage of. Again, this does not provide any insight
into the question: ‘what is low?’ Compounding our interpretation issue is the problem that there
will be inherent differences at the firm level, industry level and global level which will need to
be explored and explained.
At the firm level, differences may arise because:

Access to debt and equity markets vary.


The availability of debt at a firm level will differ between entities and between time periods.
The riskiness of the investment and proposed projects will need to be considered by
providers of finance in context.
There may be varying managerial preferences towards debt or equity and when
management change, these preferences are likely to change also.
And so forth.

At the industry level, differences may arise because:

The types of projects being funded vary and therefore so do their financing requirements.
For example, a company investing in large assets for continuing revenue generation (eg
house building, large construction projects, aeroplanes etc) will have greater and more
regular financing requirements. Both real and relative debt levels will normally be higher in
these industries as a result.
Market perceptions of the industry’s capital requirements will vary.
And so forth.

At the global level, differences may arise because:

There will be varying levels of availability for different forms of financing owing to the
economic scenario.
Global market sentiment towards future investment and financing arrangements changes
over time.
There may be government or regulatory interventions. For example, legislation over
financing levels (financial institutions being required to maintain tier 1 capital ratios, for
example) or restrictions on credit lines to small, medium or large businesses within certain
geographies.
Global and domestic money markets demand and supply levels will vary.
It might be less beneficial given (expected) rates of inflation and the opportunity costs of
other investments.
And so forth.

There are three core ratios which are most frequently undertaken which shed light on an entity’s
exposure to financing risk:

1. capital gearing ratio;


2. interest cover;
3. dividend cover.

(a) Capital gearing ratio


There are several acceptable ways to calculate a company’s level of gearing. All things being
equal, we would like to use the following definition:

However, if we only have the financial statements available, a practical and straightforward
alternative is:

Calculation difficulties

Financing risk measures market value against market value rather than book value against
book value. These figures will be markedly different, especially for a mature and profitable
company. It would be extremely rare for a company which has traded for several years to
continue to sell their shares at book value; investors will be asked to pay a premium.
Applying book values will probably add weight to debt. Prudence might tell you, however,
that this is not an altogether bad thing. Nevertheless, it is worth noting that the gearing ratio
as calculated using this method will be for guidance purposes only.
Non-current liabilities will contain balances which are not debt instruments, for example
provisions. Ideally, the only non-current liabilities you would capture would be those which
related directly to long-term debt obligations.
Current liabilities may contain some liabilities which are being used to finance the business,
and which therefore should be included in the calculation of gearing. For example, a bank
overdraft or other short-term bank loans might be plugging a financing gap. Their short
repayment terms might contribute to a clearer understanding of financing risk if included in
the calculation.
The problems associated with the definition of shareholders’ funds were covered above in
the description of ROCE. These issues also exist here.

Interpreting the gearing ratio


As stated above, the larger the gearing ratio the greater the financing risk. Ultimately, assuming
that this stays within rational bounds, investors do not mind how high financing risk rises
provided that they are adequately and appropriately compensated for it. Various prompts on how
to interpret an increase (decrease) in the ratio are provided above.
In the case of Ryanair, the gearing ratio is 54 per cent (unadjusted) or 49.6 per cent (adjusted).
This has decreased from last year, which, although being a good sign, does still prompt one to
comment that there appears to be a heavy reliance on debt financing. We would, however, expect
an airline to maintain a high gearing ratio given their operational strategies. They buy long-lived
assets (over 20 years) and it would be poor financial management to finance these through short-
term loan obligations. An airline which has a strong purchase and early replacement policy
regarding their fleet of planes – as Ryanair appears to have – would have proportionately higher
financing requirements than other businesses in the same sector. An analyst would be more
concerned to see low levels of debt for an entity such as this.
Should an analyst be concerned about these debt levels? This is a much more difficult
question! Levels of liquidity and the amount of cash retained as a buffer suggest that any short-
term problems could be dealt with. The main problem when interpreting solvency ratios is that,
by default, we are speculating about medium- to long-term performance, ie forecast profits/cash
flows that will be generated by the business. Any interpretation that involves estimating future
demand (or events) is subject to error.
As with other ratios, therefore, an analyst would search for additional information to support
their evaluation. Not only do companies estimate future demand, but the airline industry as a
whole has a regulatory body who regularly seek to appraise what will happen in the coming
months and years. One would also like to know the gearing ratios of other companies in the same
sector, as well as gearing ratios for other companies in other comparable sectors, ie those
traditionally associated with high levels of borrowing.

Extract 3.6: The chicken or the egg: solvency problems versus liquidity?
‘Lack of solvency created bank’s problems’
Management at the Co-operative Bank ‘took its eye off the ball’ leading to significant internal management
issues, a banking analyst has claimed.
By Kevin White | Published May 15, 2013 | FTadviser (Financial Times online)
Speaking after the mutual plunged into crisis last week, Ian Gordon, a banking analyst for Investec, said the
Co-operative Bank’s problems were rooted around a lack of solvency.
Last week ratings agency Moody’s downgraded the bank’s creditworthiness on concerns it could need a
government bailout…The size of the hole in the balance sheet could be as high as £1.8bn, according to an
analyst note from Barclays this week. The note said Co-op bond holders may be targeted, as the mutual cannot
issue equity. The subsequent downgrade in its credit rating led to an almost immediate resignation of the
bank’s chief executive Barry Tootell, and the announcement of a strategic review of the entire business by new
group chief executive Euan Sutherland.
Mr Gordon said: ‘Unlike the liquidity problems faced by the likes of HBoS, Northern Rock, and Bradford &
Bingley in 2008, the Co-operative Bank’s problems lie in solvency, or a lack of it.’ The bank had a £113,300 fine
in January from the FSA over failure to handle 1629 PPI complaints correctly. It also wrote down £150m on
Finacle IT project (to date). He added that the price action seen last Friday, which saw bond prices fall by 25
per cent, ‘suggested that the market was pricing in the prospect of a default’. He added: ‘The bank is less
capitalized than some of its peers and as a mutual it has been allowed by the regulator to be run with less
scrutiny. ‘For the bank to be made as safe and secure as possible it needs a cash injection, but I don’t think
that will necessarily lead to a bailout. It’s more likely the Co-operative Group itself will step in.’
When asked whether the bank had a lack of solvency, a spokesman for the Co-operative Bank said: ‘We are
aware of the solvency issues and the need to improve capital ratios.’ When asked about Mr Gordon’s
comments that the bank had been run with less regulatory scrutiny, the spokesman added: ‘It is true that the
bank is not a plc but we regularly publish our accounts and engage with investors, the press and regulators,
and are subject to considerable scrutiny.’
According to its results for 2012, the banking division contributed 17.6 per cent of revenue to the group, with
its revenue of £2.21bn for 2012 dwarfed by the food retail division’s revenue of £7.44bn. It repeated an overall
operating loss of £280.5m for 2012 (compared to 2011 profit of £141.1m) and a loss before tax of £673.7m
(compared to 2011 profit of £54.2m).

(b) Interest cover


The interest cover ratio shows the number of times a company (theoretically) could have paid
their finance costs from their earnings. The ratio is calculated as follows:

This is therefore a guide to the solvency of the business. Again, as with other ratios, it is difficult
to suggest that there is an optimum level of coverage. We would suggest, however, that if the
interest cover ratio falls below 1, this is a strong signal that current debt levels are unaffordable,
ie the company cannot afford to pay its obligations even once. If, in the following months or
years, interest payments are not made, the lender might bring into effect claw-back clauses.
A high level of coverage is also not desirable. If a company could afford to pay its finance
costs, say 10 times over, then investors might question whether the company is taking
appropriate advantage of comparatively cheaper forms of financing, ie debt.
In the case of Ryanair, interest payments can be covered approximately six times. Though this
appears on the high side, it is probably a reflection of the company’s profitability rather than an
inappropriate capital financing strategy. Indeed, these levels of affordability would almost
certainly be interpreted as reassuring by an analyst, who could perceive the gearing levels as
being high and future profits volatile (the airline industry is subject to periods of significant
demand variability).
(c) Dividend cover
This is another useful measure, especially for a company where shareholders are regularly
rewarded by dividends (and not just capital growth). The dividend cover ratio shows an analyst
how many times the company could have paid the dividend out of available profits. It is
calculated as:

As with the interest cover ratio, there is no optimum position. A ratio of less than 1 would
indicate a lack of affordability and a reduced likelihood of the dividend remaining at that level. A
high level of coverage might have shareholders questioning why the dividend payment wasn’t
larger, especially if there is no evidence that profits are being reinvested in profitable activities.
Ryanair rarely pays dividends. Instead, their shareholders are rewarded through capital growth
(which can be captured by selling shares). However, cash levels grew as profits were retained
and the management took the decision to make a one-off dividend payment in 2011. The
coverage is 0.7 times. This is not particularly helpful to our analysis because of the one-off
nature of this event. It is interesting, in many ways, that the one-off dividend could almost be
covered out of one year’s profits. This might spur investors to query whether future dividend
payments are a reasonable request, especially if reserves continue to be retained rather than
invested in operations.

WORKED EXERCISE 3.6 Ryanair’s solvency ratios


Gearing ratio:

a
We have presented two non-current liabilities figures. The first assumes that all non-current liabilities should be
classified as debt, while the second assumes that only balances classified as ‘non-current maturities of debt’ are debt.
In neither have we included ‘current maturities of debt’ obligations (2012: €368.4m; 2011: €336.7m).
Investor ratios
Current and potential investors are the intended primary users of financial statements. The ratios
set out above would carry some interest to all stakeholders whereas the set of ratios that follow
are of the greatest value and interest to this primary user group. Again, there are many ratios an
investor might calculate to appraise their stance in relation to the entity under consideration, but
you will come across the following most regularly as they are also the most useful:

investor returns: dividend yield and capital yield;


earnings per share;
price to earnings ratio.
(a) Investor returns: dividend yield and capital yield
The dividend cover ratio allows analysts to interpret an entity’s dividend payment in book value
terms, ie how many times a company could have paid the dividend from available annual
accounting profits. Of more interest to shareholders is the yield on the investment they have
made. There are two components to an annual return on equity: dividends and capital growth.
These are calculated as:

The dividend yield ratio reveals to an investor the rate of (dividend) return on their investment.
Worked exercise 3.7 illustrates how investors can measure their annual returns. Worked exercise
3.8 shows Ryanair’s investor returns. The numbers might be slightly misleading as we are using
naive unadjusted information to interpret complex positions. For example, even though Ryanair
did not pay a dividend during 2012, there was a share buyback, which is a different way in which
a company can reward shareholders (which isn’t a dividend payment or capital growth). Also,
the share price on 31 March 2010 was €3.68 but on 1 April 2010 it rose to €4.00. This is a
significant change and leaves us unsure of which information to input as being more reliable and
relevant. Nevertheless, a basic interpretation would be that buying shares in Ryanair (as with all
airlines) could lead to highly volatile returns. During good times, the pay-off is large; in bad
times, the returns are low. If you are investing for the long term, you would need to look further
at average returns. In these periods, the dividend compensates for the negative capital growth in
2011, while the capital growth compensates for the lack of dividend in 2012.

WORKED EXERCISE 3.7

Table 3.5

Share price at 1.1.X0 Share price at 31.12.X0 Dividend for the year
Mirage plc 160p 180p 12p
Lost Illusions plc 100p 105p 3p

The dividend yield for the two entities is as follows:


Mirage plc

Lost Illusions plc


Note, however, that there is also capital growth to factor into your returns analysis.
Mirage plc

Lost Illusions plc

Therefore, the total annual return on each of these two stocks can be measured as:

Mirage plc = 20.0% (7.5% + 12.5%)


Lost Illusions plc = 8.0% (3.0% + 5.0%)

(b) Earnings per share (EPS)


The EPS is normally shown on the face of the income statement, ie it is a publicly disclosed
figure. To avoid manipulation of this figure, there are accounting regulations which govern the
calculation of this figure (IAS 33 Earnings Per Share). In summary, the basic calculation is as
follows:

The standard interpretation of EPS is that as the figure increases, the better the position of the
shareholder. Owing to the complexities of the underlying accounting and possible transactions, it
is possible (although not straightforward) for management to play around with this ratio so that it
tells investors what they want to hear. Note, however, that any accounting-led manipulation
strategy can only work for a finite period of time before any gaming is unravelled!

Expert view 3.5: IAS 33 definition of basic EPS


The lengthier explanation as set out in IAS 33 is: basic EPS is calculated by dividing profit or loss attributable to
ordinary equity holders of the parent entity (the numerator) by the weighted average number of ordinary shares
outstanding (the denominator) during the period. [IAS 33.10] The earnings numerators (profit or loss from
continuing operations and net profit or loss) used for the calculation should be after deducting all expenses,
including taxes, minority interests and preference dividends. [IAS 33.12] The denominator (number of shares)
is calculated by adjusting the shares in issue at the beginning of the period by the number of shares bought
back or issued during the period, multiplied by a time-weighting factor. IAS 33 includes guidance on appropriate
recognition dates for shares issued in various circumstances. [IAS 33.20–21]

Ryanair’s EPS is increasing. The entity’s average share capital outstanding has not moved
significantly, therefore the key driver of this change must be increased earnings. Indeed, this is
the case, as shown by the horizontal analysis as well as the profitability ratio calculations
reported above.

(c) Price to earnings (P/E) ratio


The P/E ratio is a common ratio undertaken to appraise an investment through the eyes of the
capital provider. Interpreting this ratio, however, is not straightforward. Applying an
unsophisticated view, the underlying principle is that the result gives you a multiplier, ie how
many times you would need to multiply current earnings before arriving at the price you would
pay for the share on the open market. In turn, this allows you to appraise the value of your
investment and any future investments the company intends to make, thus allowing you to price
future projects and the entity as a whole. It also provides an indication of how quickly an
investment in the stock of this entity might pay back. In other words, a P/E ratio of, say, 10 could
be interpreted to mean that the current price is 10 times the value of current earnings; or it would
take 10 consecutive earnings at that level to repay the investment at that price.
The P/E ratio should be calculated as follows:

When investors see a profitable business carrying a high P/E ratio, the expectation is for future
growth (in earnings) above inflation. The basic tenets of supply and demand intermixed with
efficient markets presumptions lead rational investors to infer that a high P/E ratio driven by a
change in the numerator, ie share price (market value per share), indicates stronger than average
demand for the stock. Where there is demand, the share price rises and this precedes the
realization of the amount paid by shareholders for this future growth. Demand is created by
above average performance, or more commonly, the promise of future above average
performance.
As you can imagine, the share price of a company can move for many reasons and accounting
earnings might not be a genuine reflection of position and current or future performance.
Therefore, not only would you need to undertake further investigation into the underlying data;
ultimately, you would also need to appraise the P/E ratio in context. Comparing P/E ratios
between companies in the same industry can be especially revealing (see online content at:
www.koganpage.com/accountingfm).
Ryanair’s P/E ratio appears high and is an indicator that the market believes that the company
will continue to be profitable over and above costs and inflation. The decline in the 2012 P/E
ratio needs to be investigated in more detail. Note that the EPS and the share price were
significantly lower in 2011 than in 2012 but the net effect is that the P/E ratio was higher. This is
because the current year (2012) EPS result did not rise as far as the current year share price.

WORKED EXERCISE 3.8 Ryanair investor ratios


Note: the share price at close 31.03.2012 was €4.48; 31.03.2011: €3.36; 31.03.2010: €3.68.
Dividend yield:
Weaknesses and limitations
As we have progressed through this chapter, the major weaknesses and limitations of the various
financial analysis techniques might have been obvious to you. Some of the questions you might
have been asking yourself are the following.
1. Is true comparability possible?
Sub-questions include:
• What entities are we comparing, and why have we chosen to compare them?
• What are the differences between these entities strategically, economically, politically,
culturally?
• Do these entities have similar accounting policies? In other words, do the financial
statements capture comparable information?
– Though accountants are required to choose policies that ensure financial information is
presented in a representation that is ‘true and fair’, managerial judgement is still required.
For example, the difference between two identical companies estimating the useful
economic lives of identical assets differently can have a serious impact on profits. If one
airline chooses to depreciate planes over 25 years while another believes they will last for
only 10 years, this will have a significant impact on accounting profit from year to year,
with the latter entity recognizing the costs earlier than the former.
– Annual results can be manipulated without needing to employ accounting flexibility.
For example, cutting back on certain items of expenditure – such as staff training and
development costs, advertising and marketing expenses – during the year will produce an
immediate improvement in profits. The medium- to long-term impact on profits,
however, will inevitably be detrimental.
– There are complex accounting transactions which can be hidden, often by posting the
movement through equity and holding back the realization of gains/losses to future
periods.
• Is comparability between time periods possible?
– Have there been one-off events during the year which have impacted
positively/negatively on the profitability of the entity?
– If so, should we (can we) adjust for these in our analysis?
– Is the broader economic/political/social/cultural climate comparable?
2. Financial statements are backwards-looking and are composed from historic information.
The preparation and audit of the annual report is a lengthy exercise and for public listed
companies, the filing deadline is three months after the year-end. Therefore, by the time you
are able to analyse the financial statements for the year ended 31 March 2012, for example,
the first quarter of the following year will be nearly complete. Regardless of business,
industry, geography or any other externality, three months can be a long time!
3. The financial statements do not tell the whole story (see Chapter 4). During the past couple
of decades we have witnessed an exponential growth in the means and methods a company
is able to use to communicate with investors and other stakeholders. The hunger from the
public for news has also made companies more proactive in their communication strategies.
Though the annual report continues to be important, its previous status as the cornerstone to
an investor’s understanding of the entity’s position and performance has been replaced. For
analysts, working with live information at their fingertips, the annual report has taken on a
more confirmatory role.
Conclusion
In this chapter, we have outlined four possible methods to analyse financial information:

horizontal analysis;
trend analysis;
vertical analysis;
ratio analysis (profitability; efficiency; liquidity; solvency; and investor ratios).

We have presented these in an accounting context and have applied each method to the annual
report. We have argued that, to extract the greatest value from your analysis, you need to view
the investment through the eyes of the interested party. For example, if you are undertaking
analysis on behalf of a potential investor who is considering buying shares in a company, you
need to understand that you are offering advice to someone who will be acquiring a stake in a
business. This business is in competition with others and therefore they also need to be
examined. The context – economic, social, political and technological – is also critically
important as it will inform your analysis.
The mechanics of financial statement analysis are relatively simple. The subsequent
interpretation of the results, however, is far more difficult. Let us presume that you are
calculating the gross profit ratio and using this information to interpret performance; one
approach you could take would be as follows:

Perform calculations: period comparisons; inter-industry comparisons.


Appraise relative performance in naive terms, ie:
– Has the ratio increased or decreased; industry ratio increased or decreased?
– State that an upward move would generally be considered positive, while downward
would be negative.
– State that a gross profit ratio higher than those of peers would be generally considered a
positive sign.
Try to explain why there has been a change:
– Has anything changed related to the business or their accounting policies comparing this
year to those previous OR between this entity and other entities in the industry OR between
the current political/economic/social/technological climate and that of prior periods?
– Use horizontal analysis, trend analysis and vertical analysis to make sense of the
accounting numbers that underpin this change (ie revenue and cost of sales).
Try to find other supporting evidence to substantiate your reasons, eg press releases.
Answer the question ‘so what?’:
– What are the implications and consequences of this change?
– Do you believe it is sustainable (or will be sustained)?
– What are the implications for the investment?
What are the limitations in the analysis you have performed?
– What further information would you need before you were satisfied that your
recommendation was appropriate and accurate?

Financial analysis is subjective. There is a wealth of data available, both quantitative and
qualitative. Part of the skill in undertaking these exercises is knowing when and where to stop.
You will need to apply judgement and this is fine-tuned through practice and experience.

COMPREHENSION QUESTIONS
1. List THREE profitability ratios and show how they should be calculated.
2. Discuss THREE possible reasons why an entity’s gross profit margin might increase from one year to the next.
3. What is meant by ‘liquidity’ and why might a supplier want to assess the liquidity position of one of their customer
companies?
4. Describe the working capital cycle and calculate the length of this for a company of your choice.
5. Show how you might appraise the solvency of an entity and state what you feel might be some red-flag issues.
6. What are the limitations of ratio analysis as a form of financial analysis?

Answers on pages 151–153

Exercises
Answers on pages 154–158

Exercise 3.1
Hawk Limited (Hawk) manufacture and distribute washing machines. The board of directors (BoD) have been
concerned for some time that their share of the market has been in decline, mainly as a result of industry-wide
competition. They are also aware that to a certain extent the industry has become the victim of its own success
as washing machines have become more reliable and durable and therefore do not need replacing as
regularly. Therefore, the BoD are considering making an investment in Sparrow Limited (Sparrow).
Sparrow manufacture dishwashers. The BoD believe that the synergistic gains will include advantages over
competitors from shared technologies, a greater distribution network, an increase in skilled employees and
management, and a shared clerical and manufacturing headquarters.
Table 3.6 shows the recently issued (summarized) financial statements of Hawk and Sparrow for the year
ended 31 October 2013.

Table 3.6

Statements of comprehensive income


Hawk Sparrow
2013 2012 2013 2012
£000s £000s £000s £000s
Revenue 12,000 13,000 9,500 7,800
Cost of sales (8,000) (8,200) (4,100) (2,600)
Gross profit 4,000 4,800 5,400 5,200
Operating expenses (2,500) (2,550) (2,600) (2,700)
Finance costs (550) (500) (900) (300)
Profit before tax 950 1,750 1,900 2,200
Income tax expense (300) (600) (800) (900)
Profit for the period 650 1,150 1,100 1,300
Statements of financial position
Non-current assets 13,075 12,000 11,000 6,000
Current assets 2,600 2,800 1,500 1,400
Current liabilities (3,200) (3,000) (2,500) (1,500)
Non-current liabilities (275) (250) (4,500) (1,500)
12,200 11,550 5,500 4,400
Equity and reserves 12,200 11,550 5,500 4,400
Current assets include:
Inventories 800 900 700 400
Trade receivables 1,000 1,800 750 600
Bank 800 100 50 400
Current liabilities include:
Trade payables 1,550 1,300 650 400
Other payables 1,200 1,200 1,000 200
Current tax payable 450 500 850 900

Required:

(a) The BoD of Hawk Limited have asked you to analyse the position and performance of Sparrow Limited.
(b) As part of this analysis, the BoD have also asked that you use the information above to compare and
contrast the position and performance of Hawk Limited with Sparrow Limited. They have asked you to
conclude your analysis by stating, with reasons, whether you feel Hawk Limited should make a bid to
acquire Sparrow Limited.

Exercise 3.2
Extracts from the (summarized) financial statements of Barksdale plc, a retail group, for the year ended 31
December 2013 are shown in Table 3.7 together with an extract from the chief executive’s report that
accompanied their issue.

Table 3.7

Statement of comprehensive income


2013 2012
£m £m
Revenue 9,062 9,022
Cost of sales (5,690) (5,535)
Gross profit 3,372 3,487
Operating expenses (2,501) (2,276)
Operating profit 871 1,211
Finance income 50 65
Finance costs (215) (147)
Profit on ordinary activities before taxation 706 1,129
Income tax expense (199) (308)
Profit for the year 507 821
Statement of financial position
2013 2012
£m £m
Non-current assets 5,868 5,979
Current assets
Inventories 1,020 850
Trade receivables 45 47
Cash and cash equivalents 5 100
Other current assets 320 185
1,390 1,182
Total assets 7,258 7,161
Share capital and reserves 2,100 1,964
Non-current liabilities 2,851 3,208
Current liabilities
Trade payables 1,410 1,250
Other current liabilities 897 739
7,258 7,161

Extract from the chief executive’s report:

The year at a glance


During the year we acted decisively to meet the challenges of the global economic downturn, taking steps to
manage costs tightly and respond quickly to the changing needs of our customers. Our adjusted profits are
down on last year to £507m. This is due in part to conditions on the High Street as well as our conscious
decision to improve our value, without compromising our quality. We have built unrivalled trust in the
Barksdale brand over the last 125 years, and will not sacrifice our core principles when times get tough.
Though we have a strong emphasis on food, furniture and general merchandise, it is clothing that is our
customers’ biggest discretionary purchase and as the UK’s leading clothing retailer, with the largest market
share, it was inevitable that demand would ease off as customers reined in their spending. Although value
market share is marginally down from 11.0 to 10.7 per cent, we have held our volume market share at 11.2
per cent. We believe this is evidence that our team are in tune with our customer base. We paid a dividend of
£300 million during 2013 compared to £350 million which was paid out in 2012.

Required:
Based solely on the information provided above, you have been asked by a potential investor to analyse the
position and performance of Barksdale plc.

Answers to comprehension questions


1. Three possible ratios include:

But there are many others, including revenue per employee, costs per square foot of factory space and
so forth.
2. There are, of course, many reasons why an entity’s gross profit margin might increase. Simply speaking,
a change in the gross profit margin relates to a positive change in one of two variables: revenue and cost
of sales. For example, maybe revenue rose during the period proportionately quicker than cost of sales,
or cost of sales decreased proportionately quicker than revenue.
Revenue could have increased as a result of:

(a) a change in sales mix;


(b) a change in product portfolio;
(c) a change in selling price.

Cost of sales might have decreased due to:

(a) a change in production mix;


(b) a change in product portfolio;
(c) a change in purchase price.

Note, however, that an increase in the gross profit margin might not always be a good outcome. The
problem with percentages is that they deal with relative terms rather than actual. It is important also to
consider further ratio analysis and performing horizontal, trend and vertical analysis.
3. In this case, liquidity refers to an entity’s short-term financial position. In simple terms: can the company
afford to continue operating to the same level for the foreseeable future?
For a supplier this could be important, especially if the entity to whom they are supplying is a key
customer and makes up a large proportion of their output uptake. If the entity could not afford to continue,
their supply chain is under threat and they will have to start looking for other outlets. They will also
consider the level of the customer’s indebtedness and start calling in unpaid invoices. In extreme
situations, they might stop supplying the customer until their liquidity position improves.
4. The working capital cycle is calculated as:

Downloading the annual report and financial statements for any listed entity is straightforward. Think of a
brand you like/buy from/are interested in and type their name into a web browser followed by the phrase
‘annual report’ or ‘investors’ and it is normally the first link you’ll be offered. They normally come in pdf
form but there is also an option with most companies to see an electronic copy online.
5. Solvency is often calculated using the gearing ratio, ie the level of debt to equity. There are several
definitions of this formula but the most commonly used is total gearing, ie non-current liabilities divided by
capital employed. The other common ratio used to work out whether a firm has over- (or under-)borrowed
(leveraged) is the interest cover ratio.
Saying that a company is over-leveraged without inter-year and inter-firm comparisons is
inappropriate. If a company had a gearing ratio in excess of the sector average or had a steeply
accelerating gearing ratio year on year, we would be concerned.
If a company’s interest cover ratio is less than 1, this could signify impending problems and solvency
issues. If a company’s interest cover ratio is greater than, say, 10, this would be a reasonable indicator
that the firm could borrow more (obviously, caveats apply about annual performance, future capital
investment, and sustainability of returns).
Maintaining an optimum level of gearing (debt to equity) is a complex process.
6. Limitations of ratio analysis: there are many limitations, including (but not restricted to) the following:
• Understanding/interpreting ratios
– Is comparability possible?
– What entities are we comparing, and why have we chosen to compare them?
– What are the differences between these entities strategically, economically, politically, culturally?
– Do these entities have similar accounting policies? In other words, do the financial statements
capture comparable information?
– Is comparability between time periods possible?
• Financial statements are backwards-looking and are composed from historic information.
• The financial statements do not tell the whole story (see Chapter 4).

Answers to exercises

Exercise 3.1

Table 3.8

Hawk Sparrow
2013 2012 2013 2012
Gross profit margin 33.3% 36.9% 56.8% 66.7%
Net profit margin 12.5% 17.3% 29.5% 32.1%
ROCE 12.0% 19.1% 28.0% 42.4%
Current ratio 0.81 0.93 0.60 0.93
Quick ratio 0.56 0.63 0.32 0.67
Trade receivables days 30.4 50.5 28.8 28.1
Inventory holding period 36.5 40.1 62.3 56.2
Trade payables days 70.7 57.9 57.9 56.2
Interest cover 2.7 4.5 3.1 8.3
Gearing 2.2% 2.1% 45.0% 25.4%

Discussion should be rewarded appropriately with a maximum of two marks being awarded for a well-
developed point and one mark being awarded for a relevant interpretation. Answers that simply state
movements in ratios will not be rewarded with any discussion marks.
Students who choose to answer the question using a SWOT (strengths, weaknesses, opportunities and
threats) analysis should be rewarded appropriately, provided they evidence their discussions with relevant
information.

Profitability
Sparrow’s turnover has increased quite markedly but this has been matched by a downturn in profitability.
There are numerous suggestions that might account for this movement; however, it is most likely that
Sparrow is aggressively chasing new business through either new markets or new products. We would not
know for certain, however, without other evidence being made available, eg strategy documents, prior-year
financial statements and so on.
Sparrow’s ROCE has fallen. This will have a negative impact on the valuation of the firm and could knock on
to the offer that is made by Hawk for a controlling stake in Sparrow. Again, we would need more information
before we could fully interpret this movement in ROCE, such as prior-year financial statements.
A net profit margin of around 30 per cent appears high. This is true particularly as we note Hawk’s net profit
margin (NPM) of 12 per cent which operates in a similar trading environment.

Liquidity
Sparrow’s position seems quite stable as evidenced by a current ratio of 0.93 in 2012 and 0.60 in 2013.
Though textbooks would suggest this should be around 2 : 1, modern businesses rarely reach these levels
owing to borrowing being relatively cheap and accessible. In light of the credit crunch this might change.
Sparrow is holding a large amount in inventory in 2013 compared to 2012. This reflects the increased level
of trading being undertaken by the firm.
Sparrow’s working capital cycle has increased by five days from 28 days to 33 days. This appears to be as a
result of other short-term payables. One suspects that these arise as a result of new business and are most
likely to be short-term warranty provisions. New products and/or new markets might give rise to new problems
and the accountants may have preferred to take a prudent position against future losses on short-term
warranties.

Solvency
Interest cover remains consistent and Sparrow appears to be financially stable and solvent for the time being.
Non-current liabilities have trebled, however. These have increased to pay for new PPE which might have
been required to facilitate the expansion in trade evidenced by the increased turnover.

Comparison with Hawk


It would be appropriate for students to state that any conclusion about investment could not be purely based on
the limited information available. Students should be rewarded for mentioning issues such as that accounting
policies might differ, accounting practice might differ and so forth.
Profit margins appear to be lower in washing machines than in dishwashers. There is no suggestion why this
might be; however, washing machines are more common household appliances than dishwashers and
therefore dishwashers might attract a margin premium based on their luxury product orientation in the market.
The declines in margins and ROCE particularly highlight that the BoD are right to consider a change of
strategy and, by engaging with a firm with higher margins, might be involving themselves in a bootstrapping
exercise. It might be possible to lift their margins, reduce their cost of sales (COS) and increase their PE ratio
by investing in an apparently less risky, higher-yield industry. The empirical proof of success for such strategies
is mixed.
Hawk appears to be liquid and solvent, judging by the working capital position and the level of debt.
At a glance it would appear that Hawk could borrow significantly more than it is currently doing and thereby
potentially adjust its financing risk and reduce its cost of capital. The gearing ratio is very low and this appears
to be the quickest route to financing the investment in Sparrow. The issue therefore is not necessarily
affordability but, instead, whether this is the right strategy.

Conclusion
The BoD are right to consider a diversification and it is logical that shareholders in a company that makes
washing machines will not object too heavily to an investment in a company that makes dishwashers.
It would appear that this is a worthwhile investment; however, this conclusion is based only on the evidence
presented. We would need to know a lot more about the quantitative and qualitative benefits of the synergistic
gains and whether Sparrow’s growth strategy will continue or whether 2013 (or 2012) was an anomaly year in
terms of results.

Exercise 3.2

Table 3.9

2013 2012
Gross profit margin 37.2% 38.6%
Net profit margin (using PBIT) 9.6% 13.4%
Dividend cover 1.7 2.3
Interest cover 4.3 8.7
Current ratio 0.60 0.59
Quick ratio 0.16 0.17
Trade receivables collection period in days 1.8 1.9
Inventories holding period 65.4 56.1
Trade payables days 90.4 82.4
Operating cycle in days (23.2) (24.5)
Gearing 57.6% 62.0%
Return on capital employed (before finance income) 17.6% 23.4%
Some comments about profitability
Margins have decreased, in particular the net profit margin. Though the chief executive claims that costs have
been controlled, it is clear that profitability is affected by them. This makes one believe that there is a high
proportion of fixed costs. This is likely given the nature of the business – ie high street retail. Rents will remain
static, as will utilities and other operating expenses. It is possible to identify that there are two areas where cut-
backs can be genuinely made: in production and the linked activities such as shipping etc (which seem unlikely
given the increase in revenue) and in wages and salaries. Both of these cuts might knock on to quality (whether
these be quality of service or quality of product) and this is something that Barksdale is trying to avoid.
The chief executive notes that the downturn has affected business. This is clear from the margins. The sales
mix might have changed, which has led to a lower GP margin, but it seems more likely that goods have had to
be discounted to sell.
What is strange, however, is that revenue has gone up. Even if we consider inflationary adjustments it
seems that there is the indication that production has not been cut back. The chief executive remarks that
people are buying less. Is this evidence that the sales mix has changed? Are fewer people buying more
expensive goods? This is speculation but seems to be a likely answer.
It might be that there is a revenue recognition problem but it is impossible to tell from what has been
provided.
The ROCE has decreased significantly. This will concern investors. The chief executive is right to try to
explain away this drop. It is worth noting, however, that the ROCE appears high, especially given the economic
downturn and the very low interest rates.
Investors will be kept ‘happy’, one suspects, with the dividend payment.

Some comments about liquidity


There are no significant changes to the current ratio or quick ratio but it is worth noting that they are both quite
low. This is not unusual for a retail business which trades with very few trade receivables, looks to maintain a
small amount of rolling stock (inventories) and will probably seek to maximize its trade payables days in terms
of lengthening its working capital cycle. Given the nature of the business and the size and scale of operations,
inventories are always likely to be reasonably high, which might put a strain on flexibility and adaptability in the
immediate short term.
On the surface, if we were simply to consider the ratios we might think that there are problems; however, if
we set these ratios in context it is less worrying.
What we might comment on is the low and diminishing cash balance and the high gearing ratio. Are more
funds available when the cash runs out? If not, how will the firm maintain its working capital given that payables
days have already been extended out to 90 days? If I were an investor I would probably have concerns over
this.

Some comments about solvency


The gearing ratio is high at over 50 per cent. See comments above about future sources of finance.
It would be important to know the maturity profile of this debt and whether facilities need to be renewed
imminently.
The interest cover figure does not suggest that this is a time for warning bells; however, it has halved in the
last 12 months. This is definitely a figure to watch, and any further negative changes might portend future
liquidity and solvency problems.
The SWOT analysis might include:

Strengths: diverse range of goods; size; economies of scale; revenue strong; well-managed operating cycle;
market share.
Weaknesses: economic climate; commodities prices; market saturation/competition.
Opportunities: overseas growth; brand regeneration; strong balance sheet for extra fundraising; consumer
demand recovery in certain sectors.
Threats: exchange rates; high-street position costly; online competition; barriers to entry might mean that the
firm is a potential takeover target.

Conclusion
On balance the company appears to have suffered owing to the economic climate and though there are
possible concerns about quality and about adaptability, it seems that the business is in reasonable financial
health and the position and performance do not raise any significant concerns.
04
Financial analysis: Part II

OBJECTIVE
The objective of this chapter is to further develop an understanding of financial analysis and interpretation skills. The first
financial analysis chapter focused upon the interpretation of financial statements as presented in the annual report. This
second chapter provides an overview of some other key forms of corporate financial communication and an introduction to
some basic underlying theory.

LEARNING OUTCOMES
After studying this chapter, the reader will be:

Able to understand that financial information is communicated through varying means and in different forms.
Aware that corporate communication may be factually accurate but it might be presented, written or distributed in a
way that sends out information signals.

KEY TOPICS COVERED


Different forms of corporate communication.
Examples and an overview of corporate communication use, misuse and abuse.

MANAGEMENT ISSUES
We continue to build on Chapter 3 which leads with the presumption that interpreting and analysing financial information
is a key skill for managers and their decision making. We move away from the financial statements and look at other parts
of the annual report as well as other forms of corporate communication.

Introduction
Despite owning the business, shareholders are not allowed access to a company’s premises
without agreed authorization, let alone being granted access to the underlying records. Instead
investors, as with all stakeholders, rely on the company to provide them with information. The
financial statements – contained within the annual report – were studied in the previous chapter
and are one means of communicating financial information. These are not, however, by any
means the only way that financial information is disclosed by an entity.
Within the annual report there are several sections which disclose information that could be
important to your financial analysis. You might also like to consider reviewing the following
when undertaking your analysis:

reflections from members of the senior management team on position, performance and
strategy;
operating and financial review/management commentary/management discussion and
analysis (title varies dependent upon jurisdiction);
directors’ remuneration report;
corporate social responsibility report;
corporate governance procedures, practices and policies;
background information on company directors.

Indeed, the annual report is not even the only place where a company produces a set of financial
statements. For large companies, interim reporting is required (quarterly or half-yearly).
Companies also regularly release earnings statements which show the current performance and
position; though these present adjusted numbers, they are driven by the IFRS-led financial
statements. Whenever significant transactions or events occur – for example a sale of new shares,
a proposed merger or acquisition – financial statements are released along with forecasts of
future results. To view financial information simply in terms of financial statements is to adopt
too narrow a perspective.
Several studies from the 1980s and 1990s show that the annual report is either the most
important or second most important source of information for users of financial statements. A
more recent study by Abraham, Marston and Darby (2012) investigating the sources of
information perceived to be most useful for risk analysis is presented in the following table:

Table 4.1 Sources of information perceived to be most useful for risk analysis (Abraham
et al, 2012)

Information source Mean Standard deviation


Meetings with management 4.5 0.7
Results announcements 4.2 0.9
Trading statements 4.1 0.8
Peer companies 4.0 0.5
Annual report & accounts 3.9 1.0
Industry experts 3.8 1.0
Analysts 3.6 1.1
Interim statements & quarterly reports 3.6 1.0
Interim reports and accounts 3.4 1.0
Market news 3.2 1.0
Newspapers 3.0 0.9
Financial news channels 2.8 0.9
Internet bulletins 2.3 1.1

Though it would be useful to discuss all of these forms of communication, there are constraints
over time and syllabus. Therefore, we propose to cover those which we think are most interesting
and thought provoking. We will look at:

corporate social responsibility reporting;


earnings announcements, conference calls and investor presentations;
media relations: press releases and newspaper coverage;
social media and internet bulletins.

The drive for information


Not only has there been a major expansion in the forms and means of corporate communication
in recent times, but there has also been a huge upsurge in demand for information from investors
and other stakeholders. This combination of increased supply and demand has created a beast
that seemingly cannot be sated. It has been said that every day individuals are exposed to around
13,000 separate corporate messages. When situations are out of the ordinary, the messages and
responses multiply.
A further effect of this new era for communication is that media outlets have adopted a more
sensationalist approach. This has been driven partly by the availability of material but also by an
economic pressure on news outlets brought about by increased competition.
If we take crisis events as extraordinary, it has been shown that the mass media’s
communication approaches, forms and strategies fundamentally differ from those which are
adopted on an average news day. Often these will involve delivering greater volume, bolder
immediacy and increasingly graphic portrayals of events; especially those moments of crisis or
catastrophe. In turn, the response to crisis from the (perceived) responsible organization(s) has
been shown to be moderated correspondingly. Crises elicit emotions from stakeholders and thus
there is an attribution effect and responsibility needs to be either taken or denied.
It has been stated within crisis communication research that honesty, openness and candour
are paramount, whilst the impact of poor, inappropriate or incomplete corporate communication
contributes to negative perceptions and value destruction. In other words, we know that the
messages coming out of companies can be deliberately – and sometimes appallingly – distorted
in order to manage consumers’ impressions. However, this approach to communication
manipulation needs to be carefully considered to avoid erosion of reputational capital.
Stakeholder management
The primary purpose of corporate communication is managing relationships with stakeholders.
These relationships are required to create, develop and foster reputational capital. It is necessary
to acknowledge that many businesses have stepped back from the neo-classical economic theory
of organization and accepted socio-economic theory in its stead. The word ‘accounting’ has as its
root ‘accountability’. Indeed, linguistically, these two words do not share a stem; they share
meaning. Though we acknowledge that the primary corporate objective is the maximization of
shareholder wealth, we also believe that businesses must also be accountable to stakeholders
beyond this group.
Earlier we established that the users of financial information were varied. Their interests
differ, their objectives are varied and the information they require from corporate disclosure
might not always be the same. They are, however, important to the continuance of the
organization as their acceptance sustains the reputational capital. Also related to this, it is
necessary for organizations to open up to other stakeholder groups beyond current and potential
investors because this develops checks and balances that help to protect the welfare of society.
Though it might not be possible to correlate ‘doing good’ directly with returns, a combination of
engagement, words and actions can foster intangible returns such as employee, customer and
supplier goodwill.
Companies and their management teams are aware, however, that there are different grades of
stakeholder, ie those that drive value and with whom it is essential the firm communicates fully
and transparently in order to protect reputational capital, and those where it is simply desirable.
Divisions have been made between stakeholders based upon their interests. A split between
interests that are economic and moral in nature has been suggested. A more detailed stakeholder
salience model has been developed which categorizes stakeholders according to their relative
salience to the organization. This might seem irrelevant to you at first glance; however, you
should note that it inevitably guides a company’s communication strategy and therefore will also
influence how you should approach the information that is presented.
Mitchell, Agle and Wood (1997) propose three underpinning attributes to salience:

power (ability to levy power upon an organization);


legitimacy (the legitimacy of the claim on the organization);
urgency (the degree to which a stakeholder can call for immediate action).

The stakeholder’s relative level(s) of power, legitimacy and urgency determines their level of
prominence. It also defines the depth and breadth of stakeholder-targeted communication. The
seven stakeholder categories (an eighth – non-stakeholder – is also recorded) are shown in Figure
4.1. These classifications are further analysed in Table 4.2.

Figure 4.1 Stakeholder typology: one, two or three attributes present. Source: Mitchell
et al, 1997, p 874
Table 4.2 Stakeholder classifications

Latent stakeholders – possess one (of the three) attributes.


Owing to the limited time and resources available to managers, it is likely that they will do very little about
stakeholders whom they believe possess only one of the three attributes. It is possible that their existence (ie the
latent stakeholders) will not be acknowledged. Note that this relationship cuts both ways, and these stakeholders are
unlikely to pay attention to the entity.
Category Attribute(s) Description
Possess power to impose their will on a firm, but by not having a legitimate relationship or
1. Dormant Power
an urgent claim, their power remains unused.
2. Possess the attribute of legitimacy, but they have no power to influence the firm and no
Legitimacy
Discretionary urgent claims.
Those with urgent claims but having neither power nor legitimacy are the ‘mosquitoes
3.
Urgency buzzing in the ears’ of managers: irksome but not dangerous, bothersome but not
Demanding
warranting more than passing management attention, if any at all.
Expectant stakeholders – possess two attributes.
Stakeholder salience will be moderate where two of the stakeholder attributes – power, legitimacy, and urgency – are
perceived by managers to be present.
In the situation where stakeholders are both powerful and legitimate, their influence in the
Power and firm is assured, since by possessing power with legitimacy, they form the ‘dominant
4. Dominant
legitimacy coalition’ in the enterprise (Cyert & March, 1963). The expectations of any stakeholders
perceived by managers to have power and legitimacy will ‘matter’ to managers.
When a stakeholder lacks legitimacy but has urgency and power, that stakeholder will be
5. Power and coercive and possibly violent, making the stakeholder ‘dangerous’, literally, to the firm.
Dangerous urgency ‘Coercion’ is suggested as a descriptor because the use of coercive power often
accompanies illegitimate status.
Dependent stakeholders lack power but have urgent legitimate claims; they depend upon
Legitimacy others (other stakeholders or the firm’s managers) for the power necessary to carry out
6.
and their will. Because power in this relationship is not reciprocal, its exercise is governed
Dependent
urgency either through the advocacy or guardianship of other stakeholders, or through the
guidance of internal management values.
Definitive stakeholders – possess all three attributes.
Stakeholder salience will be high where all three of the stakeholder attributes are perceived by managers to be
present.
Power, By definition, a stakeholder already exhibiting both power and legitimacy will be a member
7. Definitive legitimacy of a firm’s dominant coalition. When such a stakeholder’s claim is urgent, managers have
and a clear and immediate mandate to attend to and give priority to that stakeholder’s claim.
urgency
SOURCE: Adapted from Mitchell et al, 1997: 874–79.

It is unsurprising that the IASB has arrived at the verdict that the primary user group of the
annual report is current and potential investors. Other users are deemed secondary. Those falling
within this second group do not have the (immediate) power to demand actions, such as the
removal of board members or the halting of certain projects. The primary group’s interests are
mainly economic in nature and therefore there is a constant dialogue with them. Most large
businesses recognize the importance of reputational capital and therefore also have a
communication programme to engage with the secondary users and their (mainly) moral
concerns.

Corporate social responsibility reporting


Corporate social responsibility (CSR) reporting is an area which has developed rapidly in recent
years. Reports that were once seen as a joke in the corridors of banks and large organizations are
now state-of-the-art representations of the aims, goals, achievements and aspirations in the fields
of ethical, social and environmental performance and position. The demand for organizations to
behave responsibly has been consumer driven; ultimately, we all want to live in a better world.
See, for example, the volume of ‘fair-trade’ products available on the shelves in the supermarkets
or even the ethical banking alternatives now on hand.
Stakeholders whose salience was (and is) not high were asking companies to answer concerns
of a moral nature. Their replies have become thorough. In turn, there has been a recent reported
shift in perspective from definitive stakeholders – those with high salience – towards these
disclosures, perceiving them to have economic as well as moral value.
During the past decade, there has been a dramatic growth in investment in funds with a social
responsibility focus under professional management. This means that investors are deliberately
choosing to buy stock in businesses that meet social responsibility targets. There has been
commensurate demand for non-financial CSR information which cannot be obtained through the
traditional path of analysing financial statements.
CSR reports place emphasis on issues such as:

community matters;
health and safety;
diversity and human resources matters;
environmental programmes.

Large organizations commonly release a CSR report alongside their traditional annual report.
Therefore, though the disclosures are largely unregulated and voluntary, this has become part of
their year-end reporting process, thus fuzzying the border between what is voluntary and that
which is mandatory. The other primary route to disclosing CSR information is through the mass
media via the corporate website and/or press releases.
CSR reporting research suggests that corporate disclosures of policies, practices and strategies
are designed to achieve one of the following four, sometimes intertwined, objectives:

reputation risk management;


legitimacy;
image restoration;
impression management.

However, we do not want you to have the impression that CSR reporting is dominated by
deliberate manipulation and distortion. Indeed, there are genuine grounds for hope and optimism.
Researchers regularly highlight areas of good practice. It is the aim of the majority of those who
work in this field: to encourage improvements to information and its communication; to promote
fuller and more transparent disclosures; and to make suggestions to help bridge the information
asymmetry gap between stakeholders and management. We simply urge the reader to be aware
that all corporate communication is designed in a way that is intended to maximize shareholder
value and bolster reputational capital. Therefore, we would argue that stakeholders should be
alert for signs of information management, but praise the steps forward that have been made (and
are being made) in this field over such a short period of time.

Earnings announcements, conference calls and investor


presentations
The annual report has recently come to be regarded as a confirmatory (or regulatory) document,
mainly because large companies have regular (normally quarterly) results/earnings
announcements, sometimes referred to as press releases. These announcements provide
summarized (adjusted) IFRS numbers for the period under consideration. These announcements
also include a summary of the material movements during the period plus an analysis of the
position and performance. These range in complexity and in content but also commonly provide
an outline of the company strategy and the possible impact this might have on future
performance.
There has been a great deal of work undertaken investigating the use of promotional language
in these releases and the impact that has on the investor community. Though results are mixed, it
would appear that longer reports tend to quell the negative impact of unexpected earnings
(probably because they close the information asymmetry gap); tone affects investors’ reactions;
and framing can influence investors’ perceptions. Appendix C contains two examples of earnings
announcements documents: 1. IBM plc’s 1st Quarter 2013 earnings announcement; 2.
Aluminium Corporation of China Limited’s (CHALCO) 1st Quarter 2013 earnings
announcement.
Whereas CSR information is principally aimed at secondary stakeholders, this information is
intended for a select audience of primary stakeholders: the definitive stakeholders in Mitchell et
al’s (1997) framework. Both the earnings announcement documentation and the conference call
or investor presentation that follow are publicly accessible (available through the corporate
website and live broadcast over the internet), but the list of invitees is short. The communication
is directed at sell-side analysts so that they can provide updates to their sales teams and clients.
These events occasionally – although rarely – coincide with the release of the annual report.
Normally, however, the annual report follows between a couple of days and a couple of weeks
after the earnings announcement. Given that we operate in markets which interpret information
almost instantaneously, this time lag means that, in many cases, the annual report is out of date
before it is even sent to the printers.
There are those who argue that the conference call/investor presentation does not provide
incremental information over and above the accompanying earnings announcement. However,
analysts’ time is a key limiting resource and, therefore, others state that their attendance at these
events proves by default that they have some worth. All agree that improved timeliness is a
genuine benefit to this communication event. There is an argument that the company-enforced
limiting of stakeholder participation creates a two-tier system. This occurs even amongst the
definitive stakeholders themselves, as not all analysts and investors are invited to attend. We are
convinced by the conclusions of the seminal study by Frankel, Johnson and Skinner (1999, p
149) which stated: ‘we find that conference calls provide information to market participants over
and above the information contained in the corresponding press release [earnings
announcement], as evidenced by elevated return variances and trading volumes.’ Therefore, we
urge you to consider these events when undertaking your financial analysis.

Media relations: press releases and newspaper coverage


Communicating with, and through, the mass media has become central to organizations and the
investor relations role. There is evidence that links a favourable representation in the media with
improved performance. To achieve success, however, companies need to get their message out
and have it interpreted in the way they would like. Newspapers and news organizations are vital
to generating publicity. They are also important to preserving and enhancing reputational capital
and transmitting information to stakeholders who might otherwise be hard to reach (or
influence). The mass media therefore essentially acts as a conduit.
As stated above, the public are bombarded with daily corporate slogans and messages. It is
also worth remembering, however, that journalists are also bombarded with events and press
releases. How does one message get picked up by the press whilst others are left behind? There
are certain techniques and strategies that can be employed but often it is simply down to fads,
fashions, extraordinary events or luck.
In most mainstream newspapers, business news used to be limited to a single page buried
towards the middle of their publication. In recent times, there has been an upsurge in the demand
for this information and coverage has increased proportionately. The way that news outlets
interpret events impacts on corporate reputation. Companies have a new awareness that the way
the media interprets their financial information can have an impact far beyond their definitive
stakeholders. Some financial information can even provoke latent and expectant stakeholders
into action. See, for example, the reporting of the tax affairs of Starbucks, Amazon and Google
(Extract 4.1).

Extract 4.1: Media relations


Google, Amazon, Starbucks: The rise of ‘tax shaming’

By Vanessa Barford & Gerry Holt


BBC News Magazine (4 December 2012) (an extract from the news release: full story available at:
http://www.bbc.co.uk/news/magazine-20560359)

Global firms such as Starbucks, Google and Amazon have come under fire for avoiding paying tax on British
profits. There seems to be a growing culture of naming and shaming companies. But what impact does it have?
Companies have long had complicated tax structures, but a recent spate of stories has highlighted a number
of tax-avoiding firms that are not seen to be playing their part.
Starbucks, for example, had sales of £400m in the UK last year, but paid no corporation tax. It transferred
some money to a Dutch sister company in royalty payments, bought coffee beans from Switzerland and paid
high interest rates to borrow from other parts of the business. Amazon, which had sales in the UK of £3.35bn in
2011, only reported a ‘tax expense’ of £1.8m. And Google’s UK unit paid just £6m to the Treasury in 2011 on
UK turnover of £395m.
Everything these companies are doing is legal. It’s avoidance and not evasion. But the tide of public opinion
is visibly turning. Even 10 years ago news of a company minimizing its corporation tax would have been more
likely to be inside the business pages than on the front page. What changed? And is ‘shaming’ of companies
justifiable and effective?
Momentum has been growing for the last few years.
In September 2009, The Observer ran with the headline: ‘Avoiding tax robs our public services, declares
minister’. The paper reported that the government was planning to say tax is a ‘moral issue’ and that it was
‘determined to end avoidance and evasion’.
October 2010 – and the Vodafone case – saw the Daily Mail report: ‘Vodafone closes Oxford Street store at
£6bn tax protest’.

Another impact of tax shaming is that some people, such as 45-year-old self-employed businessman Mike
Buckhurst, from Manchester, boycott brands. ‘I’ve uninstalled Google Chrome and changed my search engine
on all my home computers. If I want a coffee I am now going to go to Costa, despite Starbucks being nearer to
me, and even though I buy a lot of things online, I am not using Amazon. ‘I’m sick of the “change the law”
comments, I can vote with my feet. I feel very passionate about this because at one point in my life I was a top
rate tax payer and I paid my tax in full,’ he says.

Occasionally the BBC website opens up stories to comment and these are subsequently ‘rated’
by other users. There is some debate in the online community about the type of stories that are
opened up and the parties concerned. Nevertheless, they provide a direct line into the minds of
stakeholders. We note from the story cases of individuals opting out of products offered by these
‘named and shamed’ companies. One of the most interesting things about this story to the
interested observer is the two most highly rated comments (see Extract 4.2). These individuals
draw attention to another company which is also alleged to be involved in this form of tax
management – Apple. Their comments suggest that there is media bias towards this organization.

Extract 4.2: Two most popular comments on the story: ‘Google,


Amazon, Starbucks: The rise of “tax shaming”’
Muesli3
4th December 2012–12:43

I am curious as to why in every single article it’s the same three companies being mentioned. How is it that
these companies are getting shamed much more than Apple with their 2% rate?!

production_malfunction
4th December 2012–13:07

I totally agree with the other comments on here regarding Apple. How these people are managing to get away
with only paying 2%, while at the same time avoiding media scrutiny, is beyond me.

BBC put down your iPhones and do your job please.

It has been found that companies participate in the framing of news items. Companies write press
releases in the hope that journalists will pick them up. Journalists rely on companies to write
press releases because they cannot be everywhere at once. This sometimes creates a peculiar
dialogue between unrelated – and sometimes conflictingly motivated – parties being reflected
outwards. Often the original press release is simply paraphrased and disseminated. With this in
mind, you will frequently find that press releases written by companies and made available to
journalists are written in the third person to make things simpler for the journalists to adopt.
Press releases tend to carry predominantly good news, ie news intended to create value, for
example the launch of a new product, undertaking a successful tender. Despite journalists being
told to eschew a positive linguistic turn and focus on facts, corporate communication via press
releases continues to be inherently optimistic in tone and therefore it is common for the related
news release to be influenced towards this position. Some have suggested they tend towards the
propagandaic. Interestingly, it has been found that the tone is hardened when the communication
addresses economic or financial matters. Nevertheless, it is interesting to see that almost every
study of corporate communication has picked up strong signals linked to attribution theory. In
other words, companies tend to attribute successes to ‘I’ or ‘we’ and failures to an externality.
Behind the scenes, it is common for companies to communicate formally and informally with
news outlets. They do not simply issue press releases and cross their fingers. They also brief
face-to-face through press conferences and meetings/interviews with company personnel. In
addition, companies actively monitor the effectiveness of their communication strategies.
In summary, the media is a perfectly acceptable way to substantiate your financial analysis
and to triangulate findings. Newspapers need to sell copy and therefore you should expect a level
of sensationalism, therefore in turn you should employ a healthy level of scepticism. However,
this can only go so far. If the information is inaccurate and leads to a significant level of value
destruction, it is likely that they will be held to account. Normally, newspapers will provide
information which has come from a reputable source (often the company itself) which extends
your analysis, lending weight to some arguments whilst rejecting the validity of others.

Social media and internet bulletins


As yet, nobody really knows how social media and internet bulletins will impact on corporate
communication. In the meantime, we can only comment on what has happened. The majority of
the world’s corporate household names maintain social media accounts, such as Facebook,
YouTube and Twitter. Their audiences are different, not just in terms of the demographic but also
the information needs. There are all sorts of other internet-based means that allow companies to
transmit information in a more accessible way to stakeholders, eg blogs, bulletin boards and so
forth.
We have observed three major changes as a result of this new media technology (although
there are bound to be many more): the immediacy of information availability; the reach of this
information; and the ability for users to comment on content. The advantages are that companies
can be first to frame an event and set the news agenda. However, the immediacy means that an
inappropriate lead, a lack of thoughtfulness in the communiqué or an inaccurate reflection of the
events can have serious effects.
As the technology and the behaviour of the actors develop, it will be possible to work out if
this new media is fad or fashion and here to stay. At the same time, an ethical framework will
surely emerge to govern the authors and the audience. This is essential for the long-term
sustainability of this form of communication. In the meantime, the digital landscape is a potential
goldmine of untapped information that will bolster your financial analysis with previously hard-
to-reach evidence. Equally, note that this might also be a swamp of managed corporate stories
which you’d be better off ignoring.

Conclusion
In summary, information resources have never been so accessible and up to date. Though there is
still a gap in the communication levels between those deemed definitive stakeholders (current
and potential investors) and others, the gap is closing and the richness and value of the extra
information to which they have access are diminishing and eroding.
Having undertaken your financial analysis based upon the annual report and financial
statements of an entity, you will undoubtedly be left with more questions than you have answers.
Turning to these additional sources – earnings statements, investor presentations, conference
calls, CSR reports, newspapers, corporate websites, social media – will provide you with an
almost inexhaustible secondary data set which can be used to support (or contradict) your initial
observations.
COMPREHENSION QUESTIONS

1. Discuss what is meant by the term ‘stakeholder salience’ and why it might be
important to incorporate this theory into a business’s communication strategy.
2. In terms of the accounting standard-setting process, provide examples of who might be
classified into each of the following categories and explain why you believe this to be
the case:
a dormant stakeholders;
b demanding stakeholders;
c definitive stakeholders.
3. Can you explain why analysts might view meetings with management as more
important than the annual report?
4. Can you explain why analysts might view conference calls/investor presentations as
more important than the annual report?
5. Can you explain why the annual report continues to be a useful document?
6. Describe one occasion where you have been influenced to buy a product or engage in
an activity as a direct result of an item of corporate communication, eg social media
update; and explain what that was, how it happened and outline why you think it
happened this way.

Answers to comprehension questions


1. Stakeholder salience: the salience model is a classification model used for stakeholder analysis and
stakeholder management purposes. Originally developed as a project management device, it has been
adopted by communication theorists to help understand how to respond to certain groups of users and
how to prioritize their information needs.
Stakeholder ‘salience’ is determined by the assessment of power, legitimacy and urgency:
– Power is defined as the LOA (level of authority) of each stakeholder. In a project-oriented organization
the highest LOA is having the power to authorize or stop a project.
– Urgency is the need for immediate action.
– Legitimacy is determining whether or not involvement of a stakeholder is appropriate.
The model allows the project team to determine the relative salience of any given stakeholder. The ‘level
of salience’ assists the project team in setting priority and the amount of attention that will be given to
each stakeholder.
2. Stakeholders in the standard-setting process:
(a) Dormant stakeholders – these are stakeholders with power but they lack urgency and legitimacy.
Ironically, there is strong empirical evidence that shows that analysts (who are frequently used as a proxy
for current and potential investors, ie the primary user group) tend to be dormant stakeholders.
(b) Demanding stakeholders – these are stakeholders who have urgency but lack power and legitimacy.
It might be an unfortunate truth, but a truth nonetheless, that groups such as environmental campaigners
tend to be demanding stakeholders in the standard-setting process. That is not to say their opinions are
not heard!
(c) Definitive stakeholders – these stakeholders have power, urgency and legitimacy. The large
accounting firms are often characterized as definitive stakeholders.
3. Meetings: there are several reasons why meetings might be thought to be more important than the
annual report, for example:
• They provide a chance to meet face-to-face and, to use an old cliché, ‘see into the whites of the
representative’s eyes’. There is a growing body of work which looks at the influence of behaviours,
including body language and tone.
• They provide a chance to ask questions rather than simply be the recipient of facts.
• There is a chance you might learn something new (although reporting regulations prevent the release of
private information).
• They provide an opportunity to build a relationship with the management of a firm you’re following. In
the short run this gives you something to impress your clients with; in the longer term it might provide you
with other information opportunities, eg being invited to dinners, conferences, events etc, thus bringing
the costs of otherwise expensive proprietary information down.
• And so forth…
4. Conference calls/investor presentations: many of the advantages listed above are the same. The
additional advantages are:
• You are given an abbreviated and summarized early report on the financial year by the management
and don’t have to wait for the full annual report to be released.
• You get the chance to hear the strategy outlined and explained first-hand by senior management.
• You get to hear fellow analysts’ questions. This might permit you a different interpretation of information.
5. The annual report: if other information sources contribute greater benefits in a more timely manner, why
does the annual report continue to be produced given the huge costs of preparation, audit, production
and dissemination? Indeed, this is not a straightforward question. Reasons might include:
• The annual report is independently audited and therefore grants the user an additional level of certainty
over its veracity and credibility.
• There is information in the annual report which is not (currently) provided elsewhere, eg the directors’
remuneration regulation report.
• The annual report has become a useful repository for both financial and non-financial information.
• The exercise of preparing an annual report might be deemed important from an internal perspective.
• The annual report is a permanent record, accessible at any time from anywhere in the world (assuming
the organization is a limited liability entity).
• And so forth.
6. The influence of media: answers will be personal. However, with 13,000 new items of corporate
communication bombarding us every day, it is highly unlikely that we are not compelled to act on some of
it from time to time. Identifying what, when and where is a field of study all of its own!
05
Business planning

OBJECTIVE
To provide an understanding of how budgeting fits into the business planning process, and the different approaches which
may be taken to budgeting.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Understand the role of budgeting within the business planning process.


Distinguish and evaluate different approaches to budgeting.

KEY TOPICS COVERED


Business planning and the role of budgets in that process.
The function and uses of budgets.
The budget-setting process – different approaches.
Practical budget-setting.
Flexible budgeting.
Zero-based budgeting.
Activity-based budgeting.

MANAGEMENT ISSUES
Managers should be aware of how budgeting contributes to the overall business planning process and be familiar with the
range of budgeting approaches available to them.

Introduction
In this chapter we will look at how budgeting fits into the overall framework of decision making,
planning and control within an organization. We will also look at some of the range of
approaches to budgeting and explore how different approaches are more appropriate to different
types of organization.
What is a budget? A budget can be defined as the plans of an organization expressed in
quantitative terms. It is usually detailed and sets out the planned income and expenditure of a
future period of time. Although budgets are primarily seen as being monetary, operational
budgets may also set out non-monetary elements such as stock levels and staffing requirements.
Typically a budget will be broken down into several levels, so that there will be a master
budget for the organization and then budgets for different divisions, functions or areas of the
business, all of which feed into the master budget.
The budget period will depend upon the needs of the organization. Most organizations prepare
detailed budgets for a 12-month period which corresponds with their fiscal year. They may also
prepare less detailed budgets for longer periods such as five years or ten years. Equally, many
organizations break their annual budget down into shorter periods such as quarterly or monthly.
In some cases weekly or even daily budgets may be used if that proves to be a useful
management tool.

Why budget?
Budgeting is an important part of the management process – the way in which an organization
sets goals and plans actions, allocates resources, controls and measures performance and rewards
people. Budgeting plays a central role in all of these activities. It therefore needs to be
understood and recognized as something broader than simply a set of numbers for income and
expenditure over the next 12 months.
However, despite its widespread use, budgeting is far from perfect as a management tool.
Indeed, there is debate within some circles as to how beneficial budgeting is. Some
commentators argue that in today’s uncertain and complex markets meaningful budgeting is not
possible. At the end of this chapter we explore some of the arguments against budgeting and
some of the alternatives which have been put forward.

Business planning and control: the role of budgets


A budget needs to be understood as a tool to help managers achieve organizational objectives.
First and foremost it is an important element of the planning and control functions within an
organization. Figure 5.1 illustrates how budgeting fits into this planning and control process.

Figure 5.1 The planning and control process


A major advantage of budgeting is that it forces an organization to be explicit about setting
long-term objectives and short-term goals and planning to achieve these. Without budgeting
there is a danger that managers concentrate purely on the day-to-day running of the business. A
budget helps identify the resources that are needed, and when they will be needed, so that the
right resources are in place at the right time.
The budgeting process, if implemented well, can integrate the many areas of an organization,
coordinating activities, communicating strategies, motivating staff, providing accountability and
transparency.
Budgets form an important mechanism for coordinating actions across different parts of an
organization. Each department or division within an organization, by working to their part of the
budget, will ensure goal congruence in terms of achieving overall organizational objectives.
Budgets form an important part of communication within organizations. When a budget is set
and communicated throughout the organization, this provides ‘top-down’ communication by
sending clear messages to employees within different divisions as to what is expected of them
over the next period. In turn, as budgets are monitored and performance is evaluated against
budgets, there is ‘bottom-up’ communication as this monitoring feeds back important
information to senior management as to how actual activities are unfolding in relation to the
original business plans.
Budgets can be an important motivational tool, particularly if incentives such as financial
bonuses are derived from performance against budgets. The effectiveness of budgets as a
motivational tool is dependent upon the degree to which employees ‘buy into’ the budget. This is
more likely to occur if they have a meaningful participation in the preparation of the budget.
These behavioural issues are explored in more detail later in this chapter.
Budgets also provide accountability from managers in terms of the objectives they pursue and
the results they achieve. Managers will be responsible for achieving budget targets and reporting
on their results.
A communicated budget also provides transparency about the allocation of resources within
the organization.
Although budgets are primarily forward-looking in terms of planning for the future, managers
can also use them to evaluate past performance. In the next chapter we will look at this
evaluation process in more detail.

Extract 5.1: The use of budgets in practice


A survey of 558 medium and large firms in the United States and Canada in 2010 found that 79 per cent used
budgets for control purposes (Libby and Lindsay 2010). A similar survey in 1987 found a comparable figure of
83 per cent of firms used budgets for control.

The budget-setting process


The budget-setting process within most organizations follows certain clear steps. There may be
some variation between organizations according to their exact budgeting approach, but Figure
5.2 sets out what is a typical budgeting process for most organizations.

Figure 5.2 The budget-setting process

Large organizations will have a budget committee which has overall responsibility for the
preparation of the budget. The budget committee is usually made up of managers from all levels
and divisions of the organization in order to achieve good cooperation and coordination. The
budget committee will oversee the following steps during the budget-setting process:

1. Communicate budget policy. The first step in budgeting is to communicate the budgeting
policy to all those involved in the budget-setting process. This is to ensure that everybody
fully understands the policy and to try to ensure that contributions to the budget are in line
with that policy. For example, the budgeting policy may be to increase sales by 5 per cent
over the next 12 months while at the same time increasing profit margins by 2 per cent. This
policy will form the basis of the individual elements of the budget and will inform the
process.
2. Determine restricting factors. Before the process of establishing the figures for the different
areas of the budget can start, the organization must consider what factors will restrict
performance over the budgetary period. There will always be some restricting factors. These
may, for example, be the production capacity of a factory, the amount of skilled labour
available in certain areas of activity, the quantity of raw materials available, or the size and
capacity of the market. If these restricting factors are not identified and taken into
consideration, there is a serious risk that an unachievable budget will be set.
3. Prepare the output budget. For most organizations the logical place to start the budget is
with the level of output. This is because it is usually the output which determines the
resources required throughout the organization. Output can be determined in terms of sales
or level of service provided. For organizations that produce multiple products or services,
this budget needs to be broken down into more detail to determine the relative mix of each
output element.
Some organizations, such as those in the public sector, may have fixed levels of
resources that constrain output according to funding available. However, even in such
situations, it is good practice to start with the desired level of output. Even with fixed
resources, different levels and mixes of outputs of sales or services can be achieved by
allocating resources in different ways.
4. Prepare initial budgets. Once the level of output has been determined, initial budgets for the
other areas of activity can be established. These budgets will include those for production,
purchasing, staffing, advertising and support services. Who is responsible for producing
each of these budgets will vary across different organizations and depends upon the
budgeting philosophy of the organization. The different approaches are discussed in the
section below on practical budget-setting.
5. Negotiate budgets. This is not a step which is included within the budgeting process of all
organizations. However, when an organization is structured in such a way as to give
divisional managers a high degree of autonomy, this is a necessary step. It is particularly
important if divisional managers hold budgets for which they become responsible at the
monitoring stage and those budgets include costs of services provided by other divisions.
The budget-holding manager should be given the opportunity to negotiate the level of
service and cost from the service provider. If this step is omitted, budget-holding managers
may be disincentivized by budgets they feel unable to work to.
6. Coordinate and review budgets. If individual budgets have been produced by the budget-
holding manager, and particularly if there has been a negotiation process as set out in step 5
above, it will be essential to coordinate and review individual budgets to ensure that they fit
together. One problem with delegation of budget-setting to individual managers is that self-
interest may interfere such that individual budgets do not fit together to produce a
coordinated overall budget. It is essential that there is goal congruence between individual
budgets, that is to say, that they all fit together towards achieving the overall goal of the
organization as set out in the budget policy in step 1. One role of the budget committees is
to oversee and mediate any negotiation process and to coordinate and review the overall
budget before final acceptance.
7. Final acceptance. Once individual budgets have been prepared, negotiated and coordinated,
there needs to be a final acceptance of the budget as a whole. At this stage the budget can be
finalized and will be ready for implementation.
8. Monitor and review results. The final stage in the budgeting process takes place once the
budget is being implemented. During implementation, actual performance should be
measured against budget at regular intervals. Any significant variances from budget should
be recorded and investigated, as they represent a deviation from the business plan. If
necessary, management action should be taken either to modify activities to bring them
back in line with the budget, or to modify the budget to reflect changes which may have
occurred since the budget was produced and which were not anticipated during the
budgeting process. Different organizations have different approaches towards dealing with
budget variances and these are explored in the next chapter which looks in more detail at
performance management.

Practical budget-setting
So far in this chapter we have considered the benefits of producing a budget and looked at the
overall budget-setting process. In this section we will look at some of the more detailed practical
aspects of budgeting.

Levels of budgeting
Large organizations will have several levels of budget which feed into a master budget as
illustrated in Figure 5.3.

Figure 5.3 Levels of budgeting


The master budget integrates all the budgets from the different business units of the organization.
It will be made up of both the operational budget and the financial budget.
The operational budget comprises those budgets which make up the details of the operations
of the business. These will include the sales budget, the production budget, direct labour and
purchasing and the overhead budget. These budgets feed into the budgeted income statement.
The financial budget includes the cash budget, the capital budget and the budgeted balance
sheet.

Different approaches to budget preparation


Budgeting is sometimes mistakenly seen as a purely accounting function. Although accountants
are typically involved in the coordination of the budget-setting process, and in the production of
budget-monitoring reports, the budget-setting process should involve employees at all levels
within the organization. If line managers as well as senior management are involved in setting
budgets, this will ensure greater ownership of the finalized budgets throughout the organization.
It will also enable the organization to take advantage of localized and specialist knowledge to
ensure a more accurate budget.
The sequence in which budgetary decisions are made is important, as this can have a
significant impact on the budget which emerges from the process. The budget-setting process can
be classified into two main approaches: a top-down approach or a bottom-up approach.

The top-down approach to budgeting


The top-down approach, as the name implies, involves senior management producing a master
budget and then devolving activity targets and expenditure limits down throughout the
organization. Once this overarching framework for the budget has been established, the details of
the budget, in terms of allocations to individual divisions or activities, can be discussed and
negotiated. Under this approach, any budget negotiation becomes one of how best to allocate
individual activity budgets in order to best achieve organizational goals.
The main advantage of the top-down approach to budgeting is that senior management are
able to set the budget in line with their strategic vision for the future of the organization. This
improves policy prioritization and coordination. Senior management can set demanding targets
which will strengthen fiscal discipline within the organization. Those organizations which use
the top-down approach to budgeting use the budgeting process as part of the communication
from senior management as to the strategic vision of the organization and how that is to be
implemented.
A disadvantage of the top-down approach is that it can lead to low commitment from
employees. If low-tier managers have had no involvement in the design of the budget, they are
less likely to take ownership of the budget and to show commitment towards achieving it.
However, a top-down approach does not necessarily mean that employees have no input into the
budgeting process. The allocation of resources to individual projects and activities can be left
open to discussion and line managers can still be given substantial freedom to negotiate detailed
spending within their overall budget allocation.

The bottom-up approach to budgeting


When a bottom-up budgeting approach is adopted, budget proposals are produced within each
division or section of the organization and then fed upwards. They are then compiled,
coordinated and integrated into a master budget.
A characteristic of this approach is that total expenditure tends to be determined through a
process of negotiating the details of the budget with each function manager. Because of this, the
bottom-up approach often works best in an unconstrained budgeting environment: an
environment in which individual line managers are free to put forward expenditure proposals if
these can be justified in terms of revenue generation. The approach is less suited to organizations
which have fixed expenditure limits; although it can be used in such situations, this will involve
substantially more negotiation and adjustment.
The rationale behind the bottom-up approach is greater involvement of divisional managers
and employees. If employees are involved in setting the budget they are more likely to be
committed to achieving budget targets.

Extract 5.2: Circle bring bottom-up management to the NHS


When Circle became the first private company to take control of a full-service hospital in the UK, it introduced a
new management philosophy which moved away from the traditional top-down approach found in the National
Health Service. The chief executive of Circle, Ali Parsa, whose background was in the banking industry,
believes in incentivizing staff to do things better and more efficiently and delegating power down so that they
are able to do so. Staff hold a 49.9 per cent stake in the business, and shares are awarded based on
performance. Employees are divided into clinical teams of between 50 and 100, each led by a doctor, a nurse
and an administrator. Each team has responsibility for its own budget, financial performance, and how well
patients do. Teams meet regularly to monitor performance and have the power to do things differently if they
believe they can improve the operation of the hospital. The ideology behind the model is that staff, if given both
a financial interest and the power to act, are better at identifying improvements and efficiency changes than
managers who are divorced from the day-to-day activities.

This approach also enables senior management to take advantage of the local knowledge and
specialist expertise of divisional managers who are ‘closer to the ground’. This avoids the
situation of senior management setting unrealistic budgets because they don’t have a detailed
knowledge of the activities and costs involved in achieving objectives within individual
functions.
Despite the advantages outlined above, the bottom-up approach does present a number of
challenges:

It can be an extremely time-consuming and costly process to involve a greater number of


people throughout the organization.
It makes it more difficult to ensure congruence towards organizational goals. Individual line
managers are likely to put forward budget proposals based upon local needs and wants
rather than those which contribute towards those of the organization as a whole.
It makes it much more difficult for senior management to maintain tight fiscal control.
Budget proposals from individual line managers will inevitably argue for increases in
expenditure.

There is also a risk that line managers who are allowed to set their own budgets will set
themselves undemanding targets, or that they will create ‘slack’ for themselves to avoid criticism
for not achieving targets. Within individual divisions, line managers have little incentive to
identify and propose savings that could be used to increase profitability or to finance new
initiatives. In a bottom-up budgeting system the budget negotiation process can become an
exercise in preserving existing levels of funding and attempting to obtain additional resources.
Another feature of the bottom-up approach is that it is inherently incremental in nature. This
makes significant reallocations between sectors or large restructuring of the budget unlikely.

Incremental budgeting
Incremental budgeting involves taking the previous year’s figures and adjusting them for any
known changes such as inflation, wages increases or changes in level of activity. The major
advantage of this approach to budgeting is that the budget-setter has a clear starting point based
upon previous actual performance. The disadvantage of incremental budgeting is that any
inefficiencies or wastage in budgets may be rolled forward year after year.
Incremental budgeting has had much bad press, particularly from those advocating alternative
approaches. Much of this criticism arises out of poor implementation of incremental budgeting
rather than the approach itself, and one should not be too hasty in dismissing many of the
benefits of this approach. The problem is that many organizations in practice simply take last
year’s budget and add a percentage to it to allow for inflation. This approach can justifiably be
criticized. However, incremental budgeting can still be an extremely good approach if done well.
If there are no major changes in the way a business operates, looking at last year’s figures as a
starting point for the next year makes a great deal of sense. However, these figures must be
scrutinized in detail to ensure that any necessary adjustments are made to remove unused budgets
and/or inefficiencies and to reflect known changes for the coming year.

Are the top-down and bottom-up approaches mutually


exclusive?
In practice, all budget preparation processes will have both a top-down and a bottom-up element.
Although there is a clear conceptual difference between the two approaches, all good budget-
setting will involve elements of both.
Top-down budgeting should not be seen as a tool for limiting the discretionary powers of line
managers. Nor does it eliminate the process of negotiating budgets and choosing between
competing programmes and activities. However, some top-down element is essential to provide
clarity during the process of prioritization through the focus on organizational-level goals. At the
same time, it is not feasible to impose detailed function budgets from above without having some
assurance that these will achieve their goals.
The two approaches of top-down and bottom-up do not therefore represent absolutes, but
rather differing emphasis of balance during budget-setting. The challenge is to find the right
balance which achieves all of the aims of the budget-setting process.

The basic steps of preparing a budget


In the previous section we explained that the overall master budget is made up of an operational
budget and a financial budget, each of which is composed of several functional budgets. There
may be functional budgets for several business units across the organization. This can mean a lot
of individual budgets to coordinate. In order to facilitate this coordination there is a logical
sequence to the preparation of a master budget. For commercial organizations the first step will
be to forecast sales and to prepare a sales budget. Other operational budgets will then follow
from this. Figure 5.4 illustrates a typical budget preparation sequence.

Figure 5.4 The budget preparation sequence

The following example illustrates the budget preparation steps set out in Figure 5.4. The example
is kept simple as it is intended to demonstrate clearly the main steps of budget preparation rather
than the detailed practical issues which may be involved during each stage.
WORKED EXAMPLE 5.1 Preparing a budget
Rad Co manufactures one model of vehicle radiator which is sold to just one customer, a major car manufacturer. The
radiators are incorporated into several different models of vehicle.
The following information has been compiled for the preparation of the budget for the next financial year:

1. Expected sales of radiators for the year are 180,000 units. The expected selling price per unit is $20. All sales
are made on credit.
2. The manufacture of each radiator requires 8 metres of tubing and 0.7 m2 of sheet metal. Over the coming year,
tubing is expected to cost $0.50 per metre and sheet metal is expected to cost $1.50 per square metre.
3. The manufacture of each radiator requires three stages, all of which are performed by computerized machines:

4. The production process has the following overhead costs:

5. The customer demands a very short supply time on radiators and order levels can fluctuate at short notice. Rad
Co therefore maintains an inventory of finished units. At the start of the year 8,000 units are expected to be held
in inventory. However, in order to ensure a greater buffer against fluctuations in demand, the production director
wants to increase inventory of finished radiators to 12,000 units by the end of the year.
6. Inventory of raw materials at the start of the year is expected to be as follows:

The production director plans to increase these inventory levels by 5 per cent during the year.
7. Administration and selling overheads are expected to be $500,000.
8. The balance sheet at the start of the year is expected to include the following figures:

9. The customer has been demanding greater credit terms and after some negotiation these will be increased
over the coming year such that closing trade receivables are expected to be 20 per cent of the total sales for
the year.
10. Closing trade payables are expected to be 8 per cent of the purchases for the year.
11. Planned capital expenditure for the year is $80,000.
12. Non-current assets are depreciated on a straight-line basis at a rate of 15 per cent on cost.

Using this information we will now illustrate how the budget is put together, following the steps set out in Figure 5.4.

The sales budget


The first step will be to produce the sales budget. This will detail both the physical quantity of sales and their financial
value. In the example this is a relatively straightforward task as these figures are given:

180,000 units × $20 = $3,600,000


In practice a variety of means will be used in order to establish the physical quantity of sales. These figures may be
based upon market research, they may be based upon known contracts or they may be aspirational in terms of
increased sales targets. A variety of statistical and mathematical techniques may be used in forecasting sales.
Establishing the sales price is a complex topic in its own right. Chapter 8 of this text examines pricing in detail.

The production budget


The level of production will be derived from the level of sales. Production must be budgeted at a level which meets
sales requirements and planned inventory levels. This can be derived as follows:

Units
Sales 180,000
Less: opening inventory 8,000
172,000
Add: planned closing inventory 12,000
Production required 184,000

Direct material usage budget


Material usage will be derived from the level of production established above. We have the number of units of
production needed and from the information in the example we know how much material is required for each unit. The
material usage budget will therefore be as follows:

Direct material purchases budget


Material purchasing must be sufficient to meet production needs and planned levels of inventory. In this example
there are only two direct materials to purchase (tubing and sheet metal). Once purchase quantities have been
established, these can be costed to establish a purchasing budget:

Tubing (m) Sheet metal (m2)


Production usage 1,472,000 128,800
Less: opening inventory 100,000 10,000
1,372,000 118,800
Add: closing inventory (Opening +5%) 105,000 10,500
Purchase quantity 1,477,000 129,300
Cost per unit $.50 $1.50
Purchase cost $738,500 $193,950

Machine usage budget


The machine usage budget will detail the number of hours of running time required for each machine together with
the total cost of that operation. In this example (Table 5.1) we can assume that production is within capacity, but in
practice it will be necessary to ensure that sufficient machine time is available to meet budgeted production levels.

Time (hours) Cost per hour ($) Total cost ($)


Cutting 0.10 × 184,000 = 18,400 12.00 220,800
Forming 0.13 × 184,000 = 23,920 20.00 478,400
Welding 0.15 × 184,000 = 27,600 15.00 414,000
1,113,200
Table 5.1 A comparison of private- and public-sector budgeting

Private sector Public sector


Market driven Resource constrained (ie funded by taxation)
Resources influenced by market demand Resources controlled by government through grant settlements
Reliance upon external sales Activity politically determined
Need for flexibility Fixed budgets
Profit oriented Service oriented
Single or limited number of objective(s) Multi (and often conflicting) objectives
Outputs identifiable and measurable Outputs subjective and qualitative

Fixed production overhead budget


The figures for production overheads are given in the example:

Administration and selling budget


The figure is provided in the example:

Capital expenditure budget


Capital expenditure must be budgeted to meet both the short-term and long-term capital needs of the organization.
Chapter 9 of this textbook looks at the capital expenditure decision in detail. In this example the figure is provided:

Workings for preparation of financial budgets


Before we can complete the cash budget, the income statement and the balance sheet, we need to calculate values
for closing inventory of both finished goods and raw materials and for closing trade receivables and trade payables.
From these we can derive a cost of goods sold. We also need to calculate the depreciation charge for the year.

Working 1: Raw materials closing inventory


From the production information we can calculate the value of the closing inventory:
Closing inventory of raw materials:

Tubing 105,000 m × $0.50 = $52,500


Sheet metal 10,500 m2 × $1.50 = $15,750
$68,250

Working 2: Finished goods closing inventory


In order to calculate the value of the closing finished goods inventory we need to first calculate the direct cost per unit:

Finished goods cost per unit: $ $


Direct materials:
tubing (8 × $0.50) 4.00
sheet metal (0.7 × $1.50) 1.05 5.05
Machining:
cutting (0.10 × $12.00) 1.20
forming (0.13 × $20.00) 2.60
welding (0.15 × $15.00) 2.25 6.05
Total direct cost per unit 11.10
Units in stock × 12,000
Closing stock value $133,200

Working 3: Cost of goods sold

$
Opening inventory 88,800
Production cost (184,000 × $11.10) 2,042,400
2,131,200
Less: closing inventory 133,200
Cost of goods sold 1,998,000

Working 4: Depreciation
Depreciation is charged at 15 per cent on cost:

Cost: $1,400,000 (opening balance) + $80,000 (expenditure) = $1,480,000


Depreciation charge for the year: $1,480,000 × 15% = $222,000
Accumulated depreciation: $460,000 (opening balance) + $222,000 = $682,000

Working 5: Closing trade receivables


Closing trade receivables are expected to be 20 per cent of total sales for the year:

$3,600,000 × 20% = $720,000

Working 6: Closing trade payables


Closing trade payables are expected to be 8 per cent of total purchases for the year:

Purchases from the direct material purchasing budget = $738,500 + $193,950 = $932,450 × 8% = $74,596

Cash budget
The cash budget sets out expected cash receipts and payments throughout the budget period and the impact on the
cash balance. This is an outline cash budget and most organizations will prepare a much more detailed forecast of
cash flows and cash requirements. This cash-flow planning is looked at in more detail in Chapter 7 of this text.

$
Receipts
Cash from trade receivables 3,313,700
(433,700+3,600,000–720,000)
(opening balance + sales – closing balance)
Payments
Payments to trade payables
(69,230+932,450–-74,596) 927,084
(opening balance + sales – closing balance)
Machine costs (from machine usage budget) 1,113,200
Fixed production overheads 560,000
Administration and selling costs 500,000
Capital expenditure 80,000
3,180,284
Net receipts (receipts – payments) 133,416
Add: opening cash balance 62,000
Budgeted closing cash balance 195,416

Budgeted income statement


$
Sales 3,600,000
Less: cost of sales (see working 3) 1,998,000
Gross profit 1,602,000
Fixed production overheads 560,000
Depreciation (see working 4) 222,000
Production profit 820,000
Administration and selling costs 500,000
Budgeted net profit 320,000

Budgeted balance sheets


$ $
Non-current assets
Cost (working 4) 1,480,000
Less: accumulated depreciation (working 4) 682,000
798,000
Current assets
Raw materials (working 1) 68,250
Finished inventory (working 2) 133,200
Trade receivables (working 5) 720,000
Cash at bank (from cash budget) 195,416 1,116,866
Current liabilities
Trade payables (working 6) 74,596
Total net assets 1,840,270
Financed by:
Ordinary share capital 800,000
Retained earnings (720,270+320,000) 1,040,270
1,840,270

Budgeting in different types of organization


The previous section outlined the basic budget-setting process. This is a generic process which is
relevant to most types of organization. However, there will be some variation and significant
differences between different types of organization. Some of these special considerations are set
out below.
Production businesses
The budget set out in Worked Example 5.1, although simple, is typical of that for a production
business. Many production businesses use a standard cost system to establish their production
budgets. A standard cost is the business’s estimated cost of producing a product based upon
previous experience of material usage and time taken together with up-to-date material costs,
direct labour costs and factory overhead costs. Standard costing is examined in more detail in
Chapter 6.

Service businesses
Organizations that provide services have a number of characteristics which make the budget-
setting process different from production businesses. Service-based businesses include
professionals and others who primarily sell their time and skills rather than a tangible product.
Service providers usually have a higher proportion of fixed costs than manufacturing businesses.
A very large part of those fixed costs will be employee costs, as services are provided by people
rather than manufacturing plants. Also, services cannot be stored in the same way as physical
inventory. This means that if the sale of a service is missed, that opportunity can be gone for
ever. Service businesses therefore tend to focus their budget-setting process around employee
costs and ensuring high utilization of employees.

Public-sector and not-for-profit organizations


Public-sector organizations include government and local government and public services such
as health services, police, fire service and the army. Other not-for-profit organizations will
include charities.
The aim of budgeting for a not-for-profit organization is to maximize the benefits from
expenditure given the resources available. Budget allocations should reflect current
organizational priorities and spending should be contained within sustainable levels.
A comparison of private- and public-sector budgeting is provided in Table 5.2.
Not-for-profit organizations sometimes approach budgeting in a different sequence by
preparing expenditure budgets first. However, levels of expenditure should be derived from
planned levels of activities, which are ‘outputs’ in the same way as sales are for commercial
organizations.

Merchandising businesses
If an organization is a merchandiser it buys in products to sell on, either wholesale or retail, and
undertakes no direct production itself. In this case there will be no production budget. Rather,
such organizations will have a merchandise purchasing budget.
Limitations and problems with budgeting
At the beginning of this chapter we mentioned that some commentators have criticized budgeting
as a management tool. In this final section we will conclude the examination of business
planning by discussing some of the perceived limitations and problems of the budgeting
techniques we have covered and looking at the alternatives which have emerged in recent years.

Extract 5.3: Frustrations with budgeting


A survey conducted in 2000 which questioned financial executives about their current experience with their
organizations’ budgeting revealed that 84 per cent of participants were frustrated with their organizations’
budgeting processes. (Comshare, 2000)

The dissatisfaction with current budgeting practices has resulted in practice-led developments in
two directions: some practitioners are seeking to improve the budgeting process and make it
more relevant, whereas others claim that budgeting should be abandoned altogether.

Should organizations budget?


Budgeting has been criticized by both practitioners and academics as being outdated and
unsuitable for the uncertainty involved in the rapidly changing post-industrial business
environment: budgeting is seen as consuming too much managerial time and the benefits are not
worth the cost; budgets inhibit firms from adapting to changes in a timely manner owing to their
fixed nature; budgeting is disconnected from strategy, thereby putting it out of kilter with the
competitive demands facing firms; and the use of budgets as a performance measure leads to
unreliable performance evaluation and promotes dysfunctional behaviour in employees. The
Beyond Budgeting Round Table (BBRT) is an international network of organizations that seek to
find business planning and control tools that could replace budgeting and help organizations
become more adaptive to change. The BBRT website contains a list of 10 specific criticisms of
budgeting.

Extract 5.4: The Beyond Budgeting Round Table


The Beyond Budgeting Round Table (www.bbrt.org) sets out 10 explicit criticisms of budgeting:

1. Budgeting prevents rapid response.


2. Budgeting is too detailed and expensive.
3. Budgeting is out of date within a few months.
4. Budgeting is out of kilter with the competitive environment.
5. Budgeting is divorced from strategy.
6. Budgeting stifles initiative and innovation.
7. Budgeting protects non-value-adding costs.
8. Budgeting reinforces command and control.
9. Budgeting demotivates people.
10. Budgeting encourages unethical behaviour and increases reputational risk.

We will examine some of these criticisms and other issues with budgeting in more detail below.
However, it should be understood that this criticism from the BBRT is not an attack directed at
business planning per se, but rather the management model which it perceives as lying behind
traditional budgeting approaches. It calls this the ‘command and control’ management model –
one which involves senior executives commanding and controlling the organization from a
corporate centre. Budgeting is perceived as a symptom of this management model which restricts
and constrains organizations that need to be more flexible and able to respond quickly to the
business environment.
We would like to examine three particular problems of budgeting in more detail. These are
the cost of budgeting, some negative behavioural aspects of budgeting and the problems of
budgeting in a volatile business environment.

Cost
Budgeting requires a considerable amount of time and effort which can use up and divert
valuable resources. As with any other ‘expenditure’ within a business, benefits should outweigh
the costs. If an organization feels that it is not getting value for money from its budgeting, it
needs to look at how it can increase the value of the budgets produced or how it can cut the cost
of budgeting.
The value of budgets can be increased by using them more effectively as tools for controlling
operational costs and for ensuring that organizational goals are met. Good budgets can be
particularly effective in controlling cash flow and thereby reducing the need for bank borrowing.
There are several ways in which the costs of budgeting can be reduced. One way is to use a
rolling budget (see below for details). Another way is through the effective use of
computerization which can both increase the efficiency and speed of budget preparation and
increase the usefulness of budgets by incorporating strategic tools such as sensitivity analysis
and scenario-building.
Defenders of traditional budgeting point out that although budgeting can be time-consuming
and costly, it is not as resource hungry as some of the alternatives proposed by its critics.

Behavioural aspects of budgeting


Budgeting is, at its core, a human activity and as such it is subject to all the behavioural problems
found in any area of human endeavour. The behavioural side of budgeting is a vast topic which
cannot be covered comprehensively within a textbook such as this. However, this section aims to
set out some of the key issues and arguments around the human dimension of budgeting that can
limit the accuracy and usefulness of budgets.
The first issue is commonly referred to as bounded rationality. This can be explained as the
fact that, when budgeting, we are looking to the future and cannot possibly know what will
happen. The accuracy of budgets will therefore be limited by the boundaries of our abilities to
predict the future. This problem is often compounded by the fact that budgets tend to be built on
the assumption that the future will resemble the past. A second and related problem is the fact
that budgeting often involves dealing with vast amounts of data. Even with the use of computers
to help assemble and process these data, there will be a human element in interpreting and
analysing them. Often the amount of data available is too much for people to be able to use
effectively as part of their decision making.
Another human problem with budgeting is that people often stick with trusted strategies,
particularly if the incremental approach to budgeting is used. The usefulness of budgets is
therefore limited by users’ inability to break out of habitual patterns of behaviour. Related to this
problem is a similar issue which is sometimes called satisficing behaviour. This refers to the fact
that managers will choose workable solutions to problems rather than optimal solutions.
Typically, if a solution to a problem is being sought, once a workable solution is found it will be
operationalized. It is not human nature to continue to seek alternative solutions once a workable
solution has been found.
Budget preparation can often involve dealing with conflicting objectives. The budget
negotiation process can be viewed as a game played between senior management and line
managers. Line managers, if given freedom to set their own budgets, may follow personal
objectives which conflict with those of the organization as a whole. In particular, managers will
exaggerate their need for resources and will set themselves targets which they know they can
meet. At the same time, senior management, aware of these tendencies, will attempt to restrict
resource allocation to that which is necessary to maintain anticipated activity levels. Research
has shown that where managers are able to put forward their own spending proposals, resource
requirements can be over-estimated by up to 30 per cent.
There is another gaming problem which relates to managerial behaviour once budgets have
been allocated. The hockey-stick effect refers to the tendency for budget holders to make sure
that all of their budgeted funds are spent by the end of the budget period, irrespective of whether
such spending is necessary. This is often done out of fear that an underspent budget will be cut in
the future – managers seek to ‘use’ their budget before they lose it. Managers may hold back on
spending during the year and then suddenly increase spending just before the end of the budget
period in order to bring spending up to the budget limit. This behaviour manifests itself in a
spending pattern which, if graphed, looks like a hockey stick, hence the name.

Budgeting in a volatile business environment


Opponents of budgeting argue that budgets are only useful tools in a stable business environment
in which future events can be predicted with reasonable accuracy. However, the modern post-
industrial business environment has proved to be anything but stable, being characterized by
constant change, innovation and technological development. This means, it is argued, that
budgets are often out of date and irrelevant even before they are implemented.
Defenders of budgeting point out that even in a volatile business environment there is still a
need for forecasting and planning. It could be argued that cash-flow planning becomes even
more important as business volatility increases. The beyond budgeting approach advocated by
the Beyond Budgeting Round Table and its supporters still includes cash-flow forecasts and
rolling cost forecasts. Many academics and management consultants therefore argue that the
need is not to move beyond budgeting but rather to improve budgeting.

Extract 5.5: Beyond budgeting at Statoil


Statoil, the Norwegian oil company, has eliminated traditional budgeting from the company’s management and
reporting processes. The company’s management saw budgeting as a barrier to what they wanted to achieve
as a global oil exploration company in a turbulent, dynamic and demanding business environment. The
problem with traditional budgeting, as they saw it, is that it tries to achieve too many things with one set of
figures. Traditional budgeting is used to control costs, allocate resources and set targets. By trying to achieve
all three of these objectives, traditional budgeting falls prey to manipulation and gaming from employees:
managers will set themselves undemanding targets in order to achieve bonuses; a requirement to spend within
budget stifles innovation and change in response to the business environment.
Statoil no longer uses budgets for oil exploration. Rather, under the new ‘Ambition to Action’ regime the
business aims to get the optimum cost level to maximize value. There is a differentiation between good and
bad costs. Good costs generate more income than you put in. The new approach aims to give more freedom
and responsibility to managers. There is still planning, forecasting and monitoring, and the business uses KPIs
(key performance indicators). However, the performance regime used attempts to move the management
mindset from a mechanical adherence to milestones and triggers to a holistic understanding of business
performance.

Improving business planning and budgeting


Not all critics of budgeting suggest that the process should be abandoned. There has been strong
interest in recent years, both among academics and practitioners, in improving budgets to make
them more useful for the current business environment. Key elements which have been identified
for improving the process are better communication and collaboration. In particular, rather than
budget negotiation being a process of arguing for and justifying budget allocations, it should be a
process of sharing and exploring views of the future operating environment.
Although there is an advocacy for going ‘beyond budgeting’, the mainstream of current
thinking on budgeting focuses on improving the budgeting process. This includes better
integration between strategic planning and budgeting, better feeding of business intelligence into
the budgeting process, greater inclusion and teamwork and less bureaucracy. These measures
will reduce the cost of budgeting and improve its effectiveness.
Three widely used developments of budgeting are rolling budgets, zero-based budgets and
activity-based budgets. We will therefore look at these developments in a little more detail.

Rolling budgets
A rolling budget is sometimes also called a continuous budget. This approach involves always
having a 12-month ongoing budget. Rather than set a budget for a fixed 12-month period, the
organization at the end of every month will add another month to their budget so that there is
always a 12-month budget in place. The advantage of this approach is that managers’ attention is
continuously placed on what will be happening over the next 12 months, rather than the
remaining months of a fixed-period budget. The process often also involves revising the 11-
month budget that was already set.
The rolling budget is extremely useful for organizations that have uncertain levels of activities
and need to respond by adjusting their capacity and operating levels. For example, a building
contractor may operate a rolling budget in order to ensure that labour and equipment are in place
during busy periods but that they are not idle when work is not available. This approach enables
both greater flexibility and tighter control through its frequent revision and updating. However,
the process can be resource-intensive and time-consuming as the organization is effectively
continuously producing a budget. It also requires a more flexible management approach, as
managers may find themselves working to constantly changing budgets. If not implemented and
managed well, this can lead to confusion and frustration.

Zero-based budgeting
Zero-based budgeting (ZBB) involves building the budget without reference to what happened in
the past. The idea behind ZBB is to avoid some of the pitfalls of incremental budgeting. These
include continuing existing inefficiencies and failure to re-evaluate how things are done. The
ZBB approach starts each budget afresh, rather than basing the budget on historical data from
previous periods. Managers must make a case for resources and their budget will be zero unless
they can justify the budget allocation they require. The advantage of this approach is that every
activity is questioned and has to be justified in terms of costs involved and benefits accrued.
Resources are therefore allocated according to results and needs and wasteful budget ‘slack’ is
eliminated. The approach also encourages managers to question the way resources are being
allocated and to look for alternatives.
Many organizations and particularly the health sector have seen substantial benefits from
using ZBB. It can result in an organization radically changing its cost structures, cutting
substantial amounts from overhead and support costs whilst increasing efficiency and
competitiveness. It encourages managers to be forward thinking in terms of identifying what
activities and resources will be needed to compete in future market conditions. Because this is
done on a ground-up basis rather than an incremental approach of targeting areas where costs can
be cut, managers must justify what to keep rather than what to remove. This can produce far
more substantial changes in cost and performance.
The ZBB process is therefore particularly useful for organizations which have recently
experienced substantial structural change such as an acquisition or a merger which may have left
a legacy of unnecessary overhead costs. Also, changes in the competitive environment may
create increasing pressures on costs such that an organization needs to re-examine the way it
delivers its goods or services.
The ZBB process is both complex and time-consuming. This can also make it costly such that
ultimately there has to be a payoff in terms of the costs and benefits. The process can also create
internal conflict within organizations as managers are forced to compete annually for budget
allocation. ZBB has also been criticized for focusing on short-term benefits to the detriment of
longer-term strategic development.
Although ZBB is a good idea in principle, many organizations have found that in practice it is
best combined with incremental budgeting. It can be useful to prepare a zero-based budget
periodically; to do so continually year after year offers little benefit, particularly if there are no
major changes to the way the organization is operating. Some organizations have therefore
incorporated ZBB by using it only every few years or when a major change occurs within the
organization. Between the ZBB sessions they revert to an incremental approach.
The ZBB approach became very popular in the 1970s, having originally been developed at
Texas Instruments in Dallas (Pyhrr, 1973). However, due largely to the practical problems
outlined above, many companies implemented the approach in some form and found that it did
not work for them. It has fallen out of popularity in recent years. It is, however, still used in
many areas of the public sector.

Extract 5.6: ZBB at InBev


InBev, the Belgium-based brewing company, has successfully used zero-based budgeting as a tool for cutting
unnecessary costs after acquisitions, enabling it to become the largest brewer in the world. The management
team at InBev have a reputation for ruthless efficiency in cost-cutting in new acquisitions to ensure that they
create value. They have achieved this by extending their ZBB practices to newly acquired subsidiaries,
requiring businesses to justify every expense each year.

Activity-based budgeting
One approach to improving the budgeting process which has gained popularity over the past
decade is activity-based budgeting (ABB).
The most comprehensive model of ABB has been developed by the Consortium of Advanced
Management, International (CAM-I), published in a book entitled The Closed Loop in 2004. This
model is illustrated in Figure 5.5.

Figure 5.5 Activity-based budgeting


ABB focuses on the activities of an organization rather than its departments or products. The
approach is based upon the concept that costs are driven by activities. It focuses on how
activities add value within the organization and expresses budgets in terms of activity costs. This
is in contrast to the traditional approach to budgeting which involves focusing on the input of
resources and identifying those in terms of functional areas.
The advantage of ABB is that the cost of activities within the organization are clearly
highlighted in a way that does not happen in the traditional budgeting approach. These activities
can then be linked back to the mission and strategic goals of the organization.
ABB has developed many of the ideas of ZBB and can be linked with activity-based costing
as part of a more general activity-based management approach.
This approach has been popular with many public-sector organizations such as law
enforcement and health care, as it enables such organizations to identify the costs of the
individual services they provide.

Conclusion
This chapter has explored how budgeting is an important part of the management process. We
have seen how corporate strategies can be deployed across business activities and departments
through the use of budgets. We have examined different approaches to the budget-setting process
and looked at the details of how a budget is prepared. We have also looked at budgeting within
its wider managerial context and examined some of the criticisms of budgeting practices,
together with some recent developments. This prepares the ground for the next chapter (Chapter
6) which looks at how budgets, once set, can be integrated into organizational performance
management.

COMPREHENSION QUESTIONS

1. What benefits may an organization derive from a formal budgeting process?


2. What are the elements of a master budget?
3. Explain the difference between incremental budgeting and zero-based budgeting.
4. What are the advantages and disadvantages of a ‘bottom-up’ approach to budgeting?
5. How would the budget-setting philosophy of a service provider differ from that of a
manufacturing business?
6. Discuss the potential conflict between using a budget as a motivational device and as a
means of control.
7. Identify and comment on three behavioural problems that might be experienced in a
system of budgetary control.

Answers on pages 203–205

Exercises
Answers on pages 205–208

Exercise 5.1: Selecting a suitable budget-setting approach


You have been elected onto the budget committee of a hospital and have been tasked with reviewing the
approach taken by the hospital towards budget-setting. The hospital has a relatively stable level of income
which comes from government allocations. At the same time it has a very high proportion of fixed costs which
are largely made up of salary and wage costs. The hospital engages in a wide and diverse range of activities.

Required:
Identify and discuss the factors which should be considered when selecting a suitable budget-setting approach
for the hospital.

Exercise 5.2: The objectives of budgeting


Ingram Co is a small engineering company. The company does not have a computerized accounting or
budgeting system, but rather the chief accountant manually produces a budget each year in conjunction with
the senior management team.
The managing director has expressed concern at the time and cost of producing the budget each year, and
has asked whether any short cuts could be taken.

Required:
Write a memo to the managing director which:

(a) explains the objectives of budgetary planning and control systems;


(b) identifies and explains the stages involved in the preparation of budgets;
(c) identifies ways in which the budget-setting process could be improved for Ingram Co.

Exercise 5.3: Budgetary information


Khan Co is a publishing business. The budget committee is scheduled to meet very soon to discuss plans for
next year’s budget-setting process. One item on the agenda is the sources of information that will be needed in
order to set next year’s budget.

Required:
Write a short briefing document for the budget committee which sets out the main sources of information that
should be used in setting next year’s budget.
Exercise 5.4: Budgetary style
Chin Co, an architectural services firm, uses a top-down budgeting approach. The budget is prepared by
Michael Chin, the CEO, and once finalized it is distributed to departmental managers for implementation.
Michael Chin sees the budget as an important means of improving company performance and he therefore
sets extremely demanding sales targets. Employees receive wage bonuses based upon their performance
against the budget.

Required:
Discuss the likely impact that Michael Chin’s budgeting style will have upon employee and business
performance at Chin Co.

Answers to comprehension questions


1. (a) Improved planning: budgeting forces management to plan for the future, not just in broad terms, and
also in detail.
(b) Improved control: budgets set expected levels of activity and the process of comparing actual results
against budget and investigating significant differences improves control of organizational activities.
(c) Better coordination of activities: budgeting is a means of tying together the activities of different
departments within an organization to ensure goal congruence.
(d) Better communication: budgeting provides both a top-down and a bottom-up means of
communication within an organization. Senior management communicates its operational expectations to
employees through the budget. Actual business activities throughout the organization are communicated
back to senior management through the budget-monitoring process.
(e) Motivation: the budget provides a means of assessing employee performance and can be the basis of
rewards such as bonuses or promotion within the organization.
(f) Delegation: the budgeting process allows senior management to delegate responsibility and authority
to make financial decisions to budget holders.
2. The master budget is the comprehensive summary and compilation of the individual budgets of an
organization. The master budget will usually be made up of two parts: the operational budget and the
financial budget. The operational budget sets out the details of the income-generating activities of the
organization, including revenues and expenses. The financial budget details the inflows and outflows of
cash and the assets and liabilities of the organization. For large organizations the master budget will be
broken down into different levels so that feeding into the master budget will be budgets for individual
business units and feeding into these budgets for different functions and departments within each
business unit.
3. An incremental budgeting approach involves using prior-year figures as a starting point for producing the
next year’s budget. The prior-year figures are adjusted for any known changes such as inflation, wage
increases, or changes in levels of activity to arrive at the figures used for the next year’s budget.
ZBB ignores prior-year figures and derives the figures for the next year’s budget by looking at
expected activities and estimates of their costs. Advocates of ZBB claim that this produces more
accurate budgets as prior-year inefficiencies are not rolled forward as would be the case with incremental
budgeting.
4. The main advantages of a bottom-up approach to budgeting are greater accuracy because of the local
knowledge of those involved in setting the budget, together with greater commitment to the budget from
managers who have been involved in the budget-setting process.
Disadvantages of the bottom-up approach are that it is more time-consuming and costly, coordinating
the budget and maintaining tight fiscal control is more difficult, and this approach leaves more scope for
managers to engage in ‘gaming’ practices.
5. Service providers usually have a higher proportion of fixed costs than manufacturing businesses. A very
large part of those fixed costs will be employee costs, as services are provided by people rather than
manufacturing plants. Also, services cannot be stored in the same way as physical inventory. This means
that if a sale of a service is missed, that opportunity is gone for ever. Service businesses therefore tend
to focus their budget-setting process around employee costs and ensuring high utilization of employees.
6. For a budget to be effective as a motivational device, managers must ‘buy into’ that budget: they must
believe that they can reasonably meet budget targets. This may mean setting less demanding targets
and this will be done if management wish to use the budget as a means of exerting tight fiscal control.
7. Three behavioural factors that might be experienced in a system of budgetary control might include:
(i) Budget holders may build slack into the figures to provide for an easy life, rather than operating in the
most efficient way.
(ii) Lack of ownership of the budget on the part of budget holders due to the imposition of the figures
without appropriate consultation and transparency in their calculation may lead to a lack of commitment.
(iii) Budgets can be a great motivator by setting the agenda for future performance but they can also
demotivate if they are seen as unrealistic.

Answers to exercises

Exercise 5.1: Selecting a suitable budget-setting approach


It is important to consider the culture within the hospital when deciding upon the level of participation in the
budgeting process. Any significant changes in the level of participation, either an increase or a decrease, may
produce resistance from line managers. For example, if senior medical staff are used to participating in the
budget-setting process, they may resent any reduction in their input. This may result in a lack of cooperation
and a lack of incentive to meet budget targets. On the other hand, an increase in participation for those not
used to such involvement may create anxiety and stress. Training will be required for managers who have no
prior experience of budgeting.
A top-down approach to budgeting will ensure that expenditure is maintained within the fixed income levels.
As a high proportion of the hospital’s costs are fixed and therefore uncontrollable, this also suggests a centrally
controlled top-down budget. However, a level of participation will be necessary in order to empower and
motivate staff.
For routine activities which have been conducted over several years, an incremental approach would be
most suitable. This will enable the budget committee to establish realistic ongoing operating costs and to
identify areas for potential savings.
Any new activities may benefit from a zero-based approach to budgeting. The budget will be built from the
ground up, based upon the aspirations and plans for the new activity.
When there is a large proportion of fixed costs, it can be difficult to control and monitor these. An activity-
based budgeting approach may be beneficial here. This will enable the allocation of these fixed costs to
individual activities which will allow for better benchmarking and monitoring.

Exercise 5.2: The objectives of budgeting


(a) It is important that Ingram Co continues to produce budgets, as budgeting is widely acknowledged as being
an important managerial tool for planning and control. Specifically, budgeting aids management in the
following activities:
(i) short-term and long-term planning and forecasting;
(ii) coordinating activities between different divisions of the business;
(iii) communicating plans to managers and employees;
(iv) motivating employees;
(v) controlling activities;
(vi) evaluating and rewarding performance.
(b) It is true that the budget preparation process can take a lot of time and effort. However, it is important that
budgets are prepared properly and that they accurately reflect the long-term objectives of the business in
order for them to be a useful management tool.
There are a number of steps which should be followed in order to ensure an accurate budget. These
include:
(i) communicating details of budget policy;
(ii) determining restricting factors;
(iii) initial preparation of individual operating and financial budgets;
(iv) negotiation of budgets;
(v) coordination and review;
(vi) final acceptance.
(c) An obvious improvement which Ingram Co can make is to computerize the budget-setting process. This will
both speed up the process and improve the accuracy and usefulness of budgets.
A computerized budget, produced on a spreadsheet for example, is much easier to adjust and revise
than one which has been produced manually. Not only will this help in the negotiation and coordination of
budgets, but it will also allow more sophisticated budgetary modelling such as scenario-building and ‘what
if’ analysis. Also, computerization will take the focus away from simply compiling the numbers and allow
management to focus more fully on the real planning process.
Once the budget has been computerized, it will be relatively easy to incorporate actual results and
perform variance analysis to enable ongoing monitoring of actual performance against the budget. This
also means that budgets can easily be adjusted later in the year if this is necessary.
The process of changing from manual to computerized budgets may not save time in the first year, but
in future years the process will take much less time as managers can simply take the existing spreadsheet
and adjust it for known changes.

Exercise 5.3: Budgetary information


The budget committee will need to gather information from a wide range of sources in order to ensure that the
budget for the coming year is accurate and truly reflects the strategic aims of the business. The sources of
information can be classified into two main groups: internal sources of information; and external sources of
information.
Internal sources of information will include:

1. the organization’s short-term and long-term plans;


2. previous year’s actual results;
3. operational changes from previous year;
4. staff training requirements;
5. capital expenditure plans and any non-current asset requirements.

External sources of information will include:

1. an analysis of the firm’s economic environment, including market trends and the activities of competitors;
2. estimations of inflation;
3. forecasts of exchange-rate movements for overseas sales and purchases;
4. suppliers’ price changes.

Exercise 5.4: Budgetary style


Chin Co uses a top-down budgeting approach. As operational managers are not involved in the budget-setting
process, they are less likely to take ownership of the budgets and strive to work to achieve them. This may
have a negative impact on staff morale as employees feel that their views are not being taken into account.
Furthermore, the company misses the opportunity of gaining ideas and incorporating the knowledge and
experience of operational managers into the budgets. This means that budgets are less likely to be accurate
and workable.
The difficult sales targets set by Michael Chin may be unachievable. This will have a negative impact on staff
morale and upon performance. If employees feel that they are not able to achieve their performance-related
bonuses, they may not be motivated to try to achieve targets. This can create a corporate culture of getting
away with doing the bare minimum.
If sales targets are over-estimated, related operational costs may also be too high. As a consequence, when
the projected sales are not achieved, profit margins will be reduced. The business may well be overstaffed for
its actual level of sales, which means that employees may be idle some of the time. This is both demotivating
for employees and costly to the business.
06
Budgets and performance management

OBJECTIVE
To provide an understanding of the role of budgets, standards and variance analysis in performance evaluation.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Interpret a variance report and assess its implications for management intervention.
Identify behavioural aspects of budget management.
Recommend strategies to prevent or remedy adverse behavioural aspects of budget management and to harness
positive aspects.
Evaluate alternative views on performance management.

KEY TOPICS COVERED


Standard costing and variance analysis.
Profit-related performance measures.
Performance measurement in not-for-profit organizations.
Behavioural aspects of performance management: gaming; achievement motive; creative accounting.
Alternative views on performance management: activity-based costing (ABC); balanced scorecards (BSC); just-in-
time (JIT); and total quality management (TQM).

MANAGEMENT ISSUES
The primary concern of managers is not the calculation and production of performance management data, but rather their
interpretation and analysis.
Introduction
In this chapter we will build upon your knowledge from Chapter 5 by looking at the way
feedback on actual performance is compared against budget to enable appropriate management
decisions and action.
In its broadest sense, performance management is about ensuring that the goals of the
organization are consistently met in an effective and efficient manner. It is a multidisciplinary
activity that includes aspects of human resource management, financial management, operations,
marketing and systems management and management accounting as well as strategic planning
and analysis. There are therefore many perspectives on performance management.
In order to better understand how this chapter fits into the wide-ranging activity of
performance management, it is useful to sub-categorize performance management into two broad
areas – strategic and operational – although in practice these often overlap.
Strategic performance management is concerned with implementing the strategy of the
organization and, if necessary, challenging the validity of that strategy as a means of achieving
organizational goals.
Operational performance management is concerned with managing operations to
ensure that they stay in line with the corporate strategy. An operational performance
management system can therefore be understood as a set of metrics used to quantify and measure
the activities of the organization to give feedback to managers on their actions.
It is this latter area of performance management that is the primary focus of this chapter,
although we will touch on some broader strategic issues. We will look at the sources of
management information for goal-setting and performance management and the different
performance measures which can be used, both financial and non-financial. We will also look at
practical aspects of performance management both in the private sector and in the not-for-profit
and public sectors.
The calculation of costing information for performance management can be a complex
arithmetic exercise. This information is usually produced by accountants and presented to
managers in the form of performance management reports. This chapter therefore focuses on the
interpretation and analysis of performance data rather than their production.
Attempts to measure organizational performance are only meaningful with reference to some
benchmark against which efficiency and effectiveness can be judged. It is therefore essential that
the right frame of reference is chosen. Much of the development in academic thinking and
business practice over the past few decades has been focused around establishing an appropriate
framework of reference against which performance should be measured. In this chapter we will
look at some of the arguments behind recent developments and consider multidimensional sets of
performance measures which include both internal and external measures of performance
together with financial and non-financial measures.
No matter what the approach towards performance management, there are some important
principles which must always be followed. These relate to how responsibility for managing the
budget is broken down within the organization and how individual managers are held
accountable for the way they manage performance.

Responsibility centres
The first general principle of performance management is clear delineation of responsibility.
Within large organizations, responsibility for managing the master budget is broken down into
different areas and delegated to line managers. The different budget sub-units are usually
referred to as responsibility centres. A responsibility centre is the organizational unit
(division, section, branch or geographic region) that is headed by a manager who is responsible
for its activities and results. There are four main types of responsibility centre: a cost centre; a
revenue centre; a profit centre; and an investment centre.

Cost centre
In a cost centre the manager is responsible for managing costs only. The budget is allocated to
cost centres for those areas of the business which generate costs but not income. For example, a
production manager in a factory producing shoes will be responsible for managing the cost of
production but not for any revenues earned from selling the shoes.

Revenue centre
In a revenue centre the manager will be accountable for the level of revenue earned. For
example, a book retailer with a chain of 20 bookstores will have a manager responsible for the
sales revenue generated by each store. The manager may also have responsibility for some
selling expenses, but this responsibility will be limited, as the main costs of running the store
such as property costs and staffing costs will be managed by the central office.

Profit centre
In a profit centre the manager will be responsible for both costs and revenues and therefore also
profit. The manager of a profit centre usually has more autonomy than the manager of a cost or
revenue centre, as increased costs can be incurred and justified if they result in higher profits.

Investment centre
In an investment centre the manager will be responsible not just for costs and revenues but also
for managing the level of investment required to earn the revenues. Performance will therefore
be measured not just in terms of profitability but also in terms of asset turnover or return on
capital employed (ROCE). The performance measures for investment centres are examined in
more detail later in this chapter.
The controllability principle
The second general principle of performance management is controllability. A manager should
only be held responsible for the things over which he or she has control. If line managers are to
be given responsibility for budget areas and held accountable for the performance against budget,
it is important that delegated budgets are separated into two elements: controllable and non-
controllable. Controllable elements are those which are influenced by factors over which the
manager has control, whereas non-controllable elements are those which are beyond the
manager’s control. This level of segregation is not always easy and will depend upon the
structure and systems of the organization.
For example, material prices will normally be regarded as non-controllable by a production
manager if price increases are due to external market forces. On the other hand, material prices
will be regarded as controllable if the manager has control over the timing and source of material
purchases. Material costs may exceed budget because the manager failed to purchase sufficient
materials on time and had to source extra at short notice from an alternative supplier at a higher
cost.

Profit-related performance measurement


In this section we will look at how financial performance can be measured and evaluated for cost
centres, revenue centres and profit centres. Worked Example 6.1 provides a simple illustration of
a monthly budget for a manufacturing business. We will look at how actual performance is
reported against this budget and examine some of the techniques used by managers to better
understand why actual performance has differed from budget.

WORKED EXAMPLE 6.1 Woodburn Co – performance against budget


Woodburn Co makes and sells wood-burning stoves. It has only one model – the ‘Optiburner’. The budget for
November predicts sales of 200 stoves, giving a profit for the month of $9,000. The budget is made up as follows:

Table 6.1

$
Sales revenue ($180 per unit × 200 units) 36,000
Direct materials ($10 per m2 × 3.5 m2 × 200 units) 7,000
Direct labour ($12 per hour × 1 1/4 hrs per stove × 200 units) 3,000
Variable overheads ($8 per hour × 1 1/4 hrs per stove × 200 units) 2,000
Fixed overheads ($180,000 per year ÷ 12 months) 15,000
Total expenses 27,000
Operating profit (Sales revenue – Total expenses) 9,000
At the end of November the accountant of Woodburn Co prepares the report set out in Worked
Example 6.2. This report shows the actual results against the original budget. The original budget
was for sales of 200 units whereas actual sales have been 230 units.

WORKED EXAMPLE 6.2 Fixed budget variance report for Woodburn Co

Table 6.2

Actual results Variance Original budget


$ $ $
Sales revenue 40,250 4,250 F 36,000
Direct materials 8,349 1,349 A 7,000
Direct labour 3,105 105 A 3,000
Variable overheads 2,760 760 A 2,000
Fixed overheads 16,000 1,000 A 15,000
Total expenses 30,214 3,214 A 27,000
Operating profit 10,036 1,036 F 9,000

The difference between the budgeted figure and the actual figure is reported in the middle column and is called a
variance. Where actual results are better than budget the variance is reported as ‘favourable’ (F), as is the case with
sales revenue which exceeds budget by $4,250. Where actual results are worse than budget the variance is reported
as ‘adverse’ (A), as is the case with direct material costs which exceeds budget by $1,349.
The report shows actual results against the original budget. It is therefore not surprising to find that the actual profit
differs from the budget, as actual sales were 230 units against a budget of 200 units. A budget that remains
unchanged even if the organization deviates significantly from the originally planned level of activities is called a fixed
budget.
A problem with comparing actual outcomes with the original fixed budget is that this gives limited information. If
managers are to be able to identify problems and decide on appropriate action, they must understand why actual
costs differ from the budget.
In Worked Example 6.2 actual sales have exceeded the budget by 30 units (or 15 per cent). It would therefore be
reasonable to expect costs also to be greater than budget. From the variance report we can see that this is indeed
the case. However, what we are not able to ascertain is whether these costs are still reasonable for the level of sales
actually achieved. In order to obtain this level of detail we need to be able to ‘flex’ the budget to reflect the actual level
of activity.
A flexed budget is one which has been adjusted in order to take account of any differences between actual levels
of activity and the original budgeted level of activity. For most organizations this approach has many advantages over
fixed budgeting. Accurate estimation of actual levels of activity can be difficult and flexing the budget allows for
improvement and refinement of original estimations. It can also help reveal problem areas in the original budget and
provide management with the opportunity to correct them.
Flexing the budget is particularly important if actual levels of activity become significantly different from the original
budget owing to external factors. Monitoring against a budget which is no longer relevant can produce meaningless
budget performance reports. It is far more useful to be able to compare wage costs and material costs against what
they are expected to be for actual levels of sales rather than original and no longer relevant levels of sales.
In order to flex the budget, the original (fixed) budget figures are adjusted to what they would have been for the
actual level of sales. Worked Example 6.3 shows a budget variance report after the budget has been flexed. In this
example the original (fixed) budget would have been calculated before the start of the accounting period. The flexed
budget, on the other hand, is not produced until after the accounting period, as it is only then that the actual level of
activity is known.
WORKED EXAMPLE 6.3 A flexed budget variance report for Woodburn Co

Table 6.3

Actual Flexed budget Flexed Fixed budget Original (fixed)


results variance budget variance budget
$ $ $ $ $
Sales revenue 40,250 1,150 A 41,400 5,400 F 36,000
Direct materials 8,349 299 A 8,050 1,050 A 7,000
Direct labour 3,105 –345 F 3,450 450 A 3,000
Variable
2,760 460 A 2,300 300 A 2,000
overheads
Fixed overheads 16,000 1,000 A 15,000 0 15,000
Total expenses 30,214 1,414 A 28,800 1,800 A 27,000
Operating profit 10,036 2,564 A 12,600 3,600 F 9,000

With the use of a flexed budget it is possible to analyse the total variance into two columns:

The fixed budget variance is the difference between the original fixed budget and the flexed budget. This variance
reflects the change in costs that would be expected due to the changed level of activity.
The flexed budget variance is the difference between the actual results and the flexed budget. This shows the
variance between actual costs and the costs that would be expected for the actual level of activity.

The calculation of the variances shown in Worked Example 6.3 enables managers to obtain a
more detailed analysis of the difference between actual profit and budgeted profit for a given
period of activity. When total variance is divided between a fixed budget variance and a flexed
budget variance, it is usually the flexed budget variance which is most useful to managers in
assessing organizational performance, as this variance shows the difference between actual costs
and what costs should have been for the actual level of activity.
However, even with total variance broken down into fixed budget and flexed budget variance,
there is still a limited amount of information available to managers. The usefulness of variance
analysis is greatly enhanced if the organization uses standard costing. This enables the analysis
of variances in much greater detail.

Extract 6.1: CIMA guidance on performance reporting


CIMA (the Chartered Institute of Management Accountants) produced the following guidance on performance
reporting. The ideal monthly management pack for reporting to the board should be between 10 and 20 pages
and contain the following elements:

Executive summary with a synopsis of KPIs and identifying all key issues.
Action plan specifying corrective actions and contingencies with best/worst-case scenarios.
P&L account showing period and cumulative positions with highlighted variances against budget – and
major variances. Trend analysis shown graphically.
Projected outturn recalculated on the basis of actual performance and action plans.
Profiled cash flow summarizing actual and projected receipts, payments and balances on a regular basis
to year-end.
Capital programme – analysis of progress of major capital schemes showing percentage completion,
current and projected expenditure, completion cost and timescale.
Balance sheet showing working capital position in tabular form or using performance indicators, eg
debtor and creditor days.

SOURCE: CIMA (2003)

Standard costing
A standard cost is the planned cost of producing one unit of a product, or in some cases the
planned cost of a process. This standard is usually based upon a reasonable expectation of the
time and materials involved that allows for idle time and material wastage which would be a
normal part of the way a business operates.
Budgeting and standard costing are not the same thing and it is possible to budget without the
use of standard costs. However, there is an interrelationship between the two activities in that a
budget can be seen as being made up of standard costs multiplied by the expected level of
activity. For example, if the standard labour cost of producing one unit of a product is $6 and the
budgeted level of production is 5,000 units, then the budget for production labour will be
$30,000 (5,000 × $6). Standard costs are therefore a useful tool in budget-setting and if the
business wishes to undertake a detailed analysis of actual performance against budget, it is
helpful to have standard costing information underpinning the budget.
In practice there are a number of different ways of setting standards and different
organizations use different types of standard. Standards are usually classified as being either
basic, ideal or attainable:

A basic standard is one which represents a constant standard that will not change over
time.
– This type of standard is not widely used. It is found in some manufacturing industries that
have products with long life cycles and processes which remain unchanged over many years
of production.
– By keeping the same standard over a long period the company is able to assess the
efficiency of performance across several years.
– This type of standard is not appropriate for businesses that have changing processes and
practices or fluctuating prices, as it would not give a meaningful measure of actual
performance. Also, it would be an impractical measure for a business producing non-
standardized products or services which are made to customer specifications.
An ideal standard represents the cost of producing a product in perfect conditions, with
maximum efficiency and no wastage.
– The rationale for the use of an ideal standard is to set the highest target possible and to
strive towards this, rather than being satisfied with less than optimum performance.
– In practice an ideal standard will never be achieved and therefore actual performance will
always be below standard.
– This may be demotivating for employees and would be an inappropriate standard to use if
bonuses were attached to performance evaluation.
A currently attainable standard makes allowances for normal levels of wastage and lost
time and represents the cost under normal but efficient levels of activity.
– Such a standard represents an achievable target and is therefore more appropriate for
performance measurement.
– The concept of being ‘attainable’ is not fixed and therefore an attainable standard can be
set at a relatively easy or difficult level for performance management purposes.
– In practice such standards need to be demanding enough to provide sufficient incentive
for employees to improve beyond current levels of efficiency, but not so demanding as to
prove to be a disincentive.

Whichever approach is adopted towards standard-setting, it is important that the standards used
are accurate. The method of deriving standards is therefore extremely important and will vary
across different types of organization. Some businesses use what is known as the engineering
approach which involves industry engineers and operation managers defining processes,
routines and exact material usages and costing these. This approach is likely to be used if an
ideal standard or a highly demanding attainable standard is desired. It will also be used for a new
product or process for which the company has no past actual performance data. A second
approach involves analysing actual past costs, removing unwanted inefficiencies and setting a
standard based upon this. Businesses operating in a competitive industry will often attempt to
benchmark their costs against those of competitors. They will obtain information on the actual
performance of their competitors and set standards to match or undercut these. Many
organizations use a combination of the above approaches.

Extract 6.2: Benchmarking and performance at Samsung


Samsung, the South Korean-based multinational conglomerate, arrived at its current world-leading position
through a process of benchmarking and refinement. When he succeeded his father as Samsung Group
chairman in 1987, Lee Kun-Hee set about transforming the conglomerate from a Korean competitor to a global
leader. Mr Lee insisted that the Group’s subsidiaries should measure their performance against global leaders
in their field, rather than benchmark against other Korean companies. Business units that did not measure up
to global performance, such as sugar and paper processing, were divested even though they were profitable,
because they were not capable of achieving leadership in global markets. Investment was concentrated on a
handful of businesses deemed capable of competing globally. Mr Lee also increased the autonomy of
successful businesses by eliminating cross-business subsidies and below-market transfer prices, thereby
freeing the businesses to compete more effectively in global markets.

Standard costing and variance analysis


The use of standards enables a more detailed level of variance analysis than that which we saw in
Worked Example 6.3. This is illustrated in Figure 6.1. In particular, the use of standards enables
flexed budget variances to be analysed into two key elements. These are generally referred to as
a volume variance and a rate variance.

Figure 6.1 Analysis of profit variance

A volume variance measures the efficiency with which resources are used by showing the
difference between actual usage and standard usage. This can refer to the volume of sales (sales
volume variance), usage of materials (material usage variance) or the time taken by direct labour
to complete a process (labour efficiency variance). This volume difference is multiplied by the
standard cost to show the impact upon profit.
A rate variance (which may also be called a price variance or a cost variance) measures the
difference between the actual price paid for something (eg materials or labour) and the expected
(standard) price. This difference in price is multiplied by the actual quantity used to show the
total impact upon profit. In some cases it is possible to break this rate variance down into even
more detail through the calculation of mix and yield variances.

Variance analysis
When variance analysis is performed using standard costing, it is usual to present that analysis in
the form of an operating statement or budget reconciliation report. There is no standard format
for this information. Different organizations use different styles of report and give them different
names.
Worked Example 6.4 presents a budget reconciliation report for Woodburn Co, based upon
the information given in Worked example 6.1. This report reconciles the actual profit for the
period back to the original budgeted profit and analyses this difference in terms of the variances
set out in Figure 6.1.

WORKED EXAMPLE 6.4 Budget reconciliation for Woodburn Co

Table 6.4

$ $ $
Budgeted net profit 9,000
Sales variances:
Sales margin price –1,150 A
Sales margin volume 3,600 F 2,450 F
Direct cost of variances:
Material price –759 A
Material usage 460 F –299 A
Labour rate 1,035 F
Labour efficiency –690 A 345 F
Production overhead variances:
Fixed overhead expenditure –1,000 A
Variable overhead expenditure 0
Variable overhead efficiency –460 A –1,460 A 1,036
Actual profit 10,036

The reconciliation report in Worked Example 6.4 provides a reconciliation between the original
budgeted profit of $9,000 and the actual profit of $10,036. The following analysis examines the
variances set out in that budget reconciliation in more detail.
Sales variance
The sales variance is expressed not in terms of sales price but rather sales margin, that is to say,
the profit margin from sales. The reason for this is that managers will be primarily interested in
the impact which changes in sales have upon profitability. The total sales variance is broken
down into two elements, the price variance and the volume variance.
From the sales margin price variance, the manager can see the impact upon profitability of
any changes from the budgeted price. In the Woodburn Co example the sales margin price
variance is adverse ($1,150A), which means that the actual sales price was less than the budgeted
sales price. As this is a simple example we can calculate the actual sales price: $40,250 (actual
sales revenue) ÷ 230 (actual units sold) = $175. The budgeted sales price was $180, therefore the
sales department discounted the stoves sold by $5 each. This reduction in sales price has cost the
company $1,150 in lost profit.
However, this is not the full story. If we look at the sales margin volume variance we can see
that the reduction in price has generated an increase in volume of sales with a favourable
variance of $3,600. The importance of breaking the total sales variance down into a price and
volume variance is that it enables the manager to evaluate whether the reduction in sales price
was beneficial to the company. In this case we can see that it was. The reduction in sales price
adversely affected profit by $1,150 but the resultant increase in volume increased profits by
$3,600. Therefore, the net impact was to increase profit for the month by $2,450.

Materials variance
The total materials variance is broken down into a price variance and a usage variance. The price
variance measures the cost of materials against budget and the usage variance measures the
amount of material used in comparison to what was expected for the actual level of production.
In the Woodburn Co example the material price variance is adverse ($1,150A), which means
that materials cost more than was expected, but material usage variance is favourable ($460F),
which means that fewer materials were used than would be expected for the volume of
production. In practice these two variances are often interrelated and the reason for one may lie
with the other. For example, it could be that the material price variance is adverse because better-
quality materials which cost more were purchased. In turn, this increase in quality may have
resulted in less wastage and therefore a favourable usage variance. If this was a strategy
employed by Woodburn Co, management can assess its overall success by looking at the total
materials variance. This is adverse ($299A), which would suggest that buying more expensive
materials and using less has not been beneficial to the business because this has reduced profit
for the period by $299. This is of course an over-simplification and material usage variances can
arise for other reasons, such as labour inefficiencies, faulty production or mistakes that require
products to be scrapped. Managers always need to investigate variances carefully to understand
why they have occurred.
If multiple materials are used in a manufacturing process and it is possible to vary their mix,
for example by substituting one material for another, then the material usage variance can be
further analysed into mix and yield variances. The material mix variance measures the financial
impact of varying the mix of materials. If a more expensive material has been used rather than a
cheaper one, or the mix between the two has changed, then the overall cost of materials will be
higher. The material yield variance measures the efficiency of turning inputs into outputs. A
certain quantity of materials input should produce a certain quantity of outputs. If the mix of
materials is changed, the quantity of output may be adversely affected and this will show an
adverse material yield variance, even if the change in mix shows a favourable variance.

Labour variance
The total labour variance is divided into a rate variance and an efficiency variance. The rate
variance provides information about the hourly cost of the labour (ie the rate at which labour was
paid) and the efficiency variance assesses how efficient labour was in comparison to the standard
time it should take to build a stove. In this case the labour rate variance is favourable ($1,035F)
and the labour efficiency budget is adverse ($690A). This may suggest that lower-paid and
perhaps less experienced labour was employed and because of their lack of experience they have
been less efficient. The overall labour variance is $345 favourable, which indicates that this
change in labour strategy has been financially beneficial to the business.
A phenomenon which should be taken account of when assessing labour variances is the
learning curve. If employees are involved in a new task, they will become more efficient as
they learn how to do this more effectively. The learning curve is also seen with increases in
volume of production. As production increases, staff usually become more efficient so that the
cost per unit decreases. However, this increase in efficiency comes at a declining rate and will
top out at some point.
Two other important factors are idle time and wastage. Idle time will relate to the labour
efficiency variance and wastage will normally relate to the material usage variance but can be
connected to labour variances because less experienced employees are more likely to increase
wastage of materials. Levels of idle time and wastage can be categorized into ‘normal’ and
‘abnormal’ and it is important that a performance management system is able to differentiate
between the two.
Normal idle time or wastage will be due to the way a production process operates and so will
be expected and should be built into efficiency measures. For example, employees may be idle
during the recalibration or refitting of machinery. This does not mean that managers should not
be concerned about reducing ‘normal’ idle time or wastage, but this is a matter of improving or
re-engineering systems and procedures, rather than one of employee performance management.
Abnormal idle time or wastage, on the other hand, is not expected as part of a normal production
process and should be investigated as a matter of employee performance management.

Fixed overhead expenditure variance


Fixed overheads, by definition, are not expected to change with levels of activity. Therefore, any
variance in fixed overhead costs will be due entirely to changes in level of expenditure which
should not have occurred just because of changes in production volume. In the Woodburn Co
example there is an adverse fixed overhead variance of $1,000. This should be investigated as
there is no reason for fixed overhead costs to increase just because of the higher level of sales.

Variable overhead expenditure variance


It is normal to allocate variable overhead costs based upon labour hours. As a result, any
variances in direct labour efficiency (as seen in the direct labour efficiency variance) will also
impact upon variable overhead costs. Because of this, the total variance is broken down into an
efficiency variance and an expenditure variance. In the example it can be seen that the variable
overhead variance of $460A is entirely due to the adverse labour efficiency already identified
and not to any change in actual expenditure.
This analysis illustrates how variances on their own do not provide answers for managers, but
rather they direct managers towards asking the right questions. Variance analysis can be seen as
an example of management by exception. That is to say, by focusing on variances,
management’s attention is directed to those areas of the business that are not performing
according to plan. Many regard this as an efficient management approach.

Extract 6.3: Standard costing in practice


KPMG, the international accountancy firm, in association with CIMA (the Chartered Institute of Management
Accountants), carried out a global survey into the use of standard costing in 2010. Their report found
widespread use of standard costing but highlighted serious shortcomings in the standard costing information
used in many large manufacturing organizations. Some businesses were using outdated standards and
standards which contained uncontrollable costs in their performance management.
It was found that the best companies use more than one type of standard to support different areas of
decision making. Effective performance management focused on variance analysis and remedial action in
controllable areas of performance. In addition, in order to deal with economic volatility these businesses
undertook frequent updates of their standards to keep them relevant and useful for performance management.

Exercises: now attempt Exercise 6.1 on page 239

Performance management in investment centres


The manager of an investment centre is not just responsible for the costs, revenues and profits of
that centre, but also for the investments made in order to earn those profits. Investment centres
therefore need performance measures which address the efficiency of investments, that is to say
they address both income and assets. In this section we will look at the most commonly used
investment centre performance indicators.

Return on investment (ROI)


The return on investment (ROI), which is synonymous with ROCE, is the most commonly used
performance measure for evaluating investment centre financial performance. The ROI measures
the level of income earned in relation to the assets employed to earn that income. This is usually
expressed in the following formula:

Expert view 6.1: ROI


ROI is also used to evaluate potential investments in the form of accounting rate of return (ARR) and so is
discussed in more detail in Chapter 9. The difference is that in investment appraisal the technique is applied to
estimates of future income, but for performance measurement it is applied to historic income, ie that which has
already been earned.

The popularity and widespread usage of ROI can be attributed to the fact that, because it gives a
measure as a percentage, it allows for easy comparison of performance between different
businesses, business divisions and the same business division over time.
A second and related measure sometimes used in investment centres is the capital turnover.
This measures sales revenue in relation to assets and shows how efficiently the centre is using its
assets to generate sales:

This measure may be more appropriate for a division that is primarily responsible for income
generation and has little or no control over costs.

Residual income (RI) and economic value added (EVA)


Residual income presents an alternative means of measuring income against assets of an
investment centre. Whereas ROI measures performance in percentage terms, RI measures it in
absolute terms. RI is calculated with the following formula:

RI = After-tax operating income – (Cost of capital × Invested assets)

The cost of capital charge deducted from operating income is the company’s weighted average
cost of capital (WACC) multiplied by the total assets invested in the division. It represents the
cost of financing the assets of the division and hence the minimum acceptable level of income
for the division. RI is therefore a measure of the surplus or residual which the division earns over
and above the minimum required by the company’s investors.
In the early 1990s Stern Steward & Co consultants made some refinements to residual income
and termed their new measure economic value added (EVA). This measure has been widely
adopted over the past 20 years and research suggests that up to 25 per cent of businesses now use
it to evaluate divisional performance.

EVA = Adjusted after-tax operating income – (Cost of capital × Adjusted average invested
capital)

For EVA the operating income and capital invested are both adjusted to bring them closer to
approximation of equivalent cash figures. This is to remove distortions which occurred owing to
the way income and capital have to be reported for financial accounting purposes.

WORKED EXAMPLE 6.5 Investment centre performance


The following information relates to the performance of an investment centre:

Using these figures we can calculate the performance of the investment centre:

Which is the best measure: ROI or EVA?


As ROI and EVA represent two very different measures of divisional performance, the question
naturally arises as to which is the better. In practice each of the two measures has both
advantages and disadvantages:

The advantage of ROI is that it enables easy comparison between divisions and different
companies, particularly if they are of different sizes. Also, managers are usually more
comfortable dealing with percentages.
The disadvantage of ROI is that it can encourage divisional behaviour which is not in the
interest of the business as a whole. For example, if a division has a current ROI of 20 per
cent it will be reluctant to invest in a new project which offers a return of only 15 per cent,
as this new project would reduce the average ROI for the division. But if the overall ROI of
the company is 12 per cent, the new project with a return of 15 per cent represents a good
investment because it will increase the company ROI. When using the ROI measure it is
possible for the interests of the division and the business as a whole to conflict.

EVA overcomes this problem by moving focus away from percentages, but in doing so it suffers
from the disadvantage of being an absolute measure and therefore being less useful for
comparison with other divisions or businesses of a different size. A major advantage of EVA is
that because it makes a charge for the capital used, it raises managers’ awareness that capital has
a cost and that the balance sheet needs to be managed just as carefully as the income statement.
The EVA enables managers to assess a proper trade-off between the two.

Non-financial performance indicators


The usefulness of financial measures and variance analysis as a management control and
performance measurement system has been challenged by some commentators. One criticism is
that currently used financial measures arose in manufacturing businesses operating in a relatively
stable business environment and as such they offer insufficient focus on quality for modern
service-oriented businesses. They are also focused on the short term, which means that
management decisions will be directed towards short-term financial gains at the potential cost to
long-term sustainability and development. Another criticism is that the speed of change in the
modern business environment is a barrier to effective standard-setting and variance analysis.
Some commentators have gone as far as to claim that standard costing is now obsolete. Others
have suggested that it can be updated to incorporate qualitative considerations to make it more
suitable for modern businesses.
In Relevance Lost, an influential book published in 1987, the authors Johnson and Kaplan
outlined the limitations of short-term financial measures of performance and argued for the use
of more non-financial measures. Johnson and Kaplan claimed that short-term financial measures
have become less relevant in a modern business environment characterized by rapid change,
innovation and shorter product life cycles. They proposed that a range of non-financial
performance measures should be used, covering not just operations, but also marketing and
research and development.
Johnson and Kaplan encouraged the use of performance indicators that will better predict an
organization’s long-term goals rather than just short-term financial performance. This should, for
example, include measures of efficient product design, flexible production capability, quality,
delivery time and customer feedback. One result of these criticisms has been the development of
what has come to be known as the balanced scorecard.

The balanced scorecard


The balanced scorecard (BSC) was developed in the early 1990s as a performance management
tool which addresses some of the criticisms of traditional finance-focused measures discussed
above. It’s best-known proponents are Kaplan and Norton who published an article in the
Harvard Business Review in 1992 followed by a book, The Balanced Scorecard, in 1996. Since
then it has become the most widely adopted performance management framework.
The principle of the balanced scorecard is that it presents a mix of financial and non-financial
performance measures which are derived from corporate strategy and which focus on the main
activities required to implement that strategy. Kaplan and Norton proposed three non-financial
areas and one financial area to provide four ‘perspectives’ on performance (see Figure 6.2). A
company should choose a small number of performance measures (typically five or six) to reflect
performance from each of these perspectives and attach targets to each measure which indicate
expected levels of performance.

Figure 6.2 The balanced scorecard

Financial perspective
The financial perspective considers the organization’s performance from the point of view of the
shareholders. The primary concern is how well the organization is creating value for its owners.
Kaplan and Norton suggested three core areas that should drive the business strategy: revenue
(growth and mix); cost reduction; and asset utilization.

Customer perspective
This perspective considers the organization’s performance from the point of view of customers.
The organization should understand its customer profile and the market segments in which it
competes.

Internal perspective
This perspective is concerned with the internal workings of the organization and what must be
done well in order to satisfy customers. It involves identifying critical business processes.
Kaplan and Norton identified three key areas of added value: the innovation process; the
operations process; and the post-sales process.

Learning and growth perspective


This perspective focuses on continual improvement and innovation within the organization. It
involves a recognition that markets, products and processes are continually changing and that in
order to continue satisfying customers and making good financial returns the organization must
keep learning and developing. An emphasis of this perspective is continued investment in
infrastructure – systems, organizational procedures and people, in order to provide the capability
to perform and improve in the other three perspectives.

Using a balanced scorecard


The four perspectives detailed above are those initially put forward by Kaplan and Norton. In
practice, each organization should choose perspectives which it feels are most appropriate to its
corporate mission and strategic vision. It should then select performance measures and set targets
in line with these perspectives. Extract 6.4 shows an extract from the balanced scorecard of a
major electricity utility supply company.

Extract 6.4: Balanced scorecard for an electricity utility supplier

Objectives Performance measures Targets


Financial perspective

Maximize returns ROCE 15%

Profitable growth Revenue growth 12%


Leveraged asset base Asset utilization rate 90%

Manage operating costs Operating costs per customer $125

Customer perspective

Industry-leading customer
Customer satisfaction rating 90%
loyalty

Internal perspective

Percentage of revenue from deregulated products or


Business growth 10%
services

Continued public support Customer satisfaction (5-point scale) 4.5

Customer service excellence Promised delivery percentage 97%

Optimize core business Percentage rated capacity attained 90%

Learning and growth perspective

Market-driven skill Strategic skill coverage ratio 85%

Employee satisfaction Employee satisfaction rating (5-point scale) 4.5

World-class leadership Leadership effectiveness rating (5-point scale) 4.5

Development of the BSC


Since its introduction in the early 1990s the balanced scorecard has attracted some criticism. One
early criticism was that the model developed by Kaplan and Norton focused primarily on the
needs of US-based SMEs and as such was not very useful to other types of organization. This has
led to a number of variant models (with alternative perspectives or different numbers of
perspectives) aimed at being more appropriate to a broader range of organizations. A further
criticism has been the focus on shareholders to the detriment of other stakeholders. Current
thinking on business strategy emphasizes the importance of well-rounded stakeholder
management. There is also little empirical evidence to support the claim that the use of the
balanced scorecard produces better financial performance. Despite this, the balanced scorecard
and its various derivatives are still extremely popular and widely applied by commercial
businesses, non-profit organizations, schools, colleges, government bodies and the military.
Kaplan and Norton have gone on to develop their ideas since the initial introduction of the
balanced scorecard. In their latest book The Execution Premium, published in 2008, they
incorporate the balanced scorecard into a broader execution premium process (XPP), a
broader holistic system of implementing and monitoring strategy.

Exercises: now attempt Exercise 6.3 on page 240

Performance measurement in not-for-profit organizations


A not-for-profit organization (NPO) is one which does not aim to earn profits for the owners.
Rather, revenues allocated, earned or donated are used in pursuit of the organization’s goals.
NPOs typically refer to charities and public service organizations such as education, health,
police, fire service and social welfare. However, NPOs can also include museums, churches,
religious organizations, sports organizations and political organizations. Not-for-profit status is a
legal status which will vary from country to country.
NPOs face unique challenges in assessing their performance as the performance measures
used by commercial organizations are in many cases not relevant. For example, the variance
analysis examined in Worked Example 6.4 above was concerned with the impact upon
profitability of actual performance against budget. As generating profits is not a goal for NPOs,
this approach to performance measurement is not appropriate. The challenge for NPOs therefore
is to find alternative performance measures which better suit their organizational goals and aims.
Not only are NPOs different from commercial organizations in not being unified by a profit
motive (Table 6.5), they are also a heterogeneous category, having a wide range of goals which
lead to a wide range of activities. These activities and goals are often seen as intangible and
difficult to quantify. This increases the challenge of finding appropriate performance measures.

Table 6.5 The features which typically make NPOs different from commercial
businesses

There are no shareholders to whom management have a primary responsibility to provide a financial return.

Generating profits is not a goal of the organization.

NPOs often provide services or goods free of charge.

Income is received from individuals or organizations that are different from those receiving the benefits of the
organization’s work.
Those providing resources to the NPO often do not expect a financial return in the same way as private-sector
investors.

However, there are some points of commonality that allow for a general framework of
performance management. The activities of NPOs can in most cases be categorized into three
key areas: fundraising, management and programme implementation. Adequate performance
management therefore needs to include measurement of performance in each of these areas.
One performance measure which is common to most NPOs is the proportion of resources
spent on management and fundraising as opposed to carrying out programme activities. A
smaller proportion of resources spent on management and fundraising is seen as a positive
performance indicator, and this measure is frequently used by tax authorities and watchdog
organizations when rating NPO performance.
Although the proportion of total resources spent on programme activities is important, it is not
in itself an adequate performance measure. Good performance is not just about spending money.
It is possible for an NPO to spend a high proportion of its total resources on programme
activities but for it to be ineffectual in reaching its goals. Performance measurement for NPOs
must therefore focus on output and outcomes as well as inputs. It is therefore useful to clarify
these three important concepts of inputs, outputs and outcomes:

Inputs can be defined as all the resources used to carry out the organization’s mission and
implement projects and programmes. This will include financial resources (earned,
allocated or donated) and also staff and volunteer time.
Outputs are the level of services provided by the NPO. Outputs will usually be measured in
non-financial terms and will depend upon the goals of the organization. For example, an
anti-malaria charity may report the number of mosquito nets distributed; a hospital heart
surgery department may report the number of operations performed; a college may measure
the number of people attending workshops or training classes; a humanitarian charity may
report the number of people provided with shelter, food and water following a disaster.
Outcomes are the effects which services provided (outputs) have on the organization’s stated
mission. Outcome measures should focus on how well outputs are achieving organizational
goals. These measures attempt to gauge how effective the organization is being. For
example, the anti-malaria charity may report the number of cases of malaria in the area in
which it is working; the college may report the number of students entering new jobs.

There are many approaches to performance measurement in NPOs, but all are based around three
important factors often referred to as the ‘3Es’: economy, effectiveness and efficiency. This
framework provides a focus for NPO performance measures which cover inputs, outputs and
outcomes:

Economy is concerned with achieving goals within given levels of resource input. Most
NPOs operate with restricted resources and must therefore be able to achieve their goals
effectively with a limited budget. For example, a hospital with a fixed level of funding must
be able to operate within its fixed budget, without overspending, regardless of how effective
or efficient it is.
Effectiveness is concerned with achievement of outcomes and focuses upon the
organization’s goals. For example, a drug rehabilitation clinic may have a corporate mission
of reducing the number of people dependent upon illegal drugs. That clinic may treat a large
number of patients at a low cost and thereby be deemed efficient. However, if a high
proportion of those patients relapse, the clinic may be deemed to have low effectiveness in
achieving its goals.
Efficiency is concerned with the relationship between inputs and outputs. Efficiency
measures look at the amount of resources used to achieve outputs. For example, two care
homes may both provide similar levels of care for the same number of residents, but if one
does so at a considerably lower cost than the other, it can be considered to be more efficient.

NPO efficiency can be measured in each of the three key areas of activity: fundraising,
management and programme implementation. There may be external benchmarks for efficiency,
or it may be possible for an organization to track its efficiency performance over time with
reference to past data. For example, a hospital heart surgery department could calculate the cost
per operation performed. This could then be compared to similar data from other hospitals or
tracked over several years to measure the efficiency of the department.

Value for money (VFM) as a public-sector objective


In many public sector organizations the principle of the 3Es is implemented under the term
value for money (VFM). VFM can be defined as the optimal use of resources to achieve the
intended outcomes. An organization is said to be achieving high VFM when there is an optimum
balance between the 3Es: relatively low costs (economy), high productivity (efficiency) and
successful outcomes (effectiveness).

NPO performance measurement: an example


Extract 6.5 presents a balanced scorecard developed by the Kenya Red Cross Society. The
mission of the society is ‘to work with vigour and compassion through our networks and with
communities to prevent and alleviate human suffering and save lives of the most vulnerable’.
The society’s strategic vision is ‘to be the most effective, trusted and self-sustaining
humanitarian organization in Kenya’.
From this mission statement and strategic vision the society has derived the following four
perspectives: beneficiary/stakeholder; financial stewardship; business processes; and
organization capacity. Each of these perspectives has been given a small number of objectives
against which performance measures and targets have been set. These are set out below.

Extract 6.5: NPO Balanced scorecard


Behavioural aspects of performance management:
gaming and creative accounting
Performance measurement has an impact on the environment in which it operates and will
influence the behaviour of those whose performance is being measured. Deciding what to
measure, how to measure it and what targets to set will all influence employee behaviour.
Sometimes this can have unintended negative consequences.
One important consequence of the increased emphasis on performance measures is the
pressure that performance measurement puts on employees. Research shows that the increasing
demands of performance measurement systems can increase tension, frustration, resentment,
suspicion, fear and mistrust in employees. This in turn reduces employee job satisfaction,
increases absenteeism and has been shown to result in a reduction in long-term performance.
Other research has identified how employees respond to performance measures. One
consequence is that ‘what gets measured gets done’. If employees know that their performance is
being measured, they will inevitably focus on achieving those targets against which they know
they will be assessed. This may be to the detriment of other areas of performance. A further
problem is that multiple and diverse performance measures may generate tension and conflict
between performance targets. This can cause confusion and distrust of performance measures
and can prove to be demotivating.
Another behavioural risk of setting performance measures is that managers will take actions
to improve their measured scores through ‘creative accounting’ without improving underlying
performance. Many examples of such practices can be found across a wide range of industries.
For example, hospitals that are assessed on the length of waiting lists have redefined and delayed
the point at which a patient enters the waiting list. Repair and maintenance departments that are
assessed on how quickly they respond to repair requests similarly redefine and delay the point at
which a request is formally recognized and therefore when the ‘clock starts ticking’. In many
cases these changes, which are effectively a manipulation of the definition of waiting time, mean
that there is no actual improvement in performance.
A further problem can arise with benchmarking, which is a process intended to assess and
improve performance. Sometimes managers will use benchmarking to defend rather than
improve poor performance. The manager will focus on explaining why their department or
organization performed poorly against the benchmark, citing factors that make their situation
different from those against which it is being assessed. In such cases, the benchmarking exercise
leads to little or no improvement in performance.

Exercises: now attempt Exercise 6.4 on page 241

External influences on performance


All organizations, whether they be profit-seeking or not-for-profit, are experiencing increased
external scrutiny and pressure to meet the expectations of external stakeholders. Current
academic work on external financial reporting is greatly concerned with impression management
and measures of corporate social responsibility. As a consequence, performance management is
increasingly being used to assess the impact of organizational actions on stakeholders outside the
organization.
These developments are reflected in the widespread use of the balanced scorecard and other
similar models of performance management which include measures of the impact of the
organization’s performance on customer satisfaction, employee satisfaction or local community
satisfaction. This in turn has shaped the way in which performance management systems are
developed and the kind of performance measures that are used. Techniques are now widely used
to assist management in identifying and assessing those external factors which can impact on
organizational performance and which therefore need to be taken into consideration in
performance management. These include tools such as PEST analysis (Figure 6.3) and Porter’s
five forces model.

Figure 6.3 PEST analysis: a framework of focusing managers’ attention on those


external factors (political, economic, social and technological) that may impact upon
organizational performance

There has also been an increased need to interpret performance in the light of external
considerations and, in particular, societal ethical issues that may impact on business
performance. Performance measurement is now turning to factors such as carbon footprinting,
sustainable use of resources, ethical sourcing of materials, employee welfare, pollution and
recycling.

Performance management in modern business systems


Recent developments in management accounting such as target costing, life-cycle costing, just-
in-time (JIT), throughput accounting and total quality management (TQM) have changed the
nature and use of performance measures within many organizations. This raises the question of
whether traditional performance measurement accounting techniques such as standard costing
still have a place in modern business systems.
One of the common features of the techniques mentioned above is a move towards viewing
the organization more holistically as a system or flow of processes rather than a collection of
individual activities. Management attention is therefore directed towards improving overall
performance rather than optimizing the performance of individual areas in isolation. This can
lead to some conclusions about performance that are counter-intuitive to those focused on
traditional measures.
For example, throughput accounting is concerned with improving the overall efficiency of
total production processes. There is a focus on identifying and removing bottlenecks in
production systems. Businesses employing the principles of throughput accounting have often
found that overall performance is improved by allowing an increase in idle time in some areas of
the system and by substantially reducing production batch sizes. Under traditional variance
analysis idle time is seen as an evil which should be eradicated. Equally, it can be hard to grasp
how overall financial performance can be improved by cutting batch sizes, as this means a
greater number of smaller orders, more set-ups (with a resultant increase in idle time) and more
deliveries. These factors all results in higher costs. However, because throughput accounting
looks at the holistic picture, such changes can substantially reduce inventory levels which in turn
reduces the amount of money the business has tied up in inventory, together with cutting costs of
damaged, lost and obsolete inventory. Also, smaller batch sizes can result in a more flexible
approach to production which means meeting customer needs more closely and more quickly.
This can give the business an important competitive advantage. As a result of these changes,
sales volumes and production throughput can be significantly increased. This can create
economies of scale which can lower operating costs to offset the higher costs associated with
smaller batch sizes.
So these new accounting techniques together with new approaches to manufacturing have
brought new challenges for control and performance measurement. There is a change in
emphasis towards performance measures that encompass both financial and non-financial
aspects, such as measures of on-time deliveries, reduction in inventory, cooperation with
suppliers, process cost reduction, quality improvement, reduced cycle time and product
complexity.
Furthermore, these new techniques bring with them the aims of continuous innovation,
change and development. For example, a JIT approach involves continuous improvement and a
commitment to constant change. Performance measurement systems are required that encourage
employees to focus on the critical elements of efficient operations and to provide effective links
across the value chain. Continuous improvement and innovation are also important aspects of
TQM. This approach needs performance measurement systems that incorporate benchmarking
against industry competitors and the integration of quality and strategic information. Within this
context, traditional management accounting performance measures can be regarded as an
impediment. Focus has moved from recording and reporting costs and cost variances against
budget, to understanding and controlling the causes of costs. Organizations need organic and
flexible performance management systems.
However, these new developments do not totally negate the usefulness of standards and
standard costing. In many organizations they have been integrated into these new techniques. For
example, target costing, with its emphasis on continuous improvement, requires a review of
resources used in the past to identify where fewer resources might be used in the future. This can
involve reviewing old standards and setting new standards for improvement.

Extract 6.6: Performance management at Toyota


Toyota, the Japanese manufacturing giant, employs the philosophy of just in time as part of its performance
management and production control systems. The company calls its approach the Toyota Production System
(TPS). The company’s philosophy is to ‘make only what is needed, when it is needed, and in the amount
needed’. This approach means that the focus of performance management is upon the flow of parts into
production and the flow of items through the production process. Parts should be available just when they are
needed within the production process and they should be available in the right place at that time. Toyota
concentrates its performance management on measuring and fine-tuning work cycle times, workflow, optimum
movement of products and reducing waste time, materials and capacity.

Exercises: now attempt Exercise 6.2 on page 240

Conclusion
We started this chapter by describing how performance management is a multidisciplinary
activity which extends beyond accounting to virtually every area of organizational management.
We have described some of the basic principles of performance management and looked in some
detail at traditional techniques. We have also looked at recent developments in performance
management and how tools and techniques have responded to changes in modern management
practices. In particular, this chapter has emphasized how performance management has moved
away from purely financial measures to a broader perspective which encompasses non-financial
aspects of performance and the considerations of external stakeholders.

COMPREHENSION QUESTIONS

1. What are the four main ways of delegating responsibility for budget management?
2. What are the advantages and disadvantages of using flexible budgets?
3. What type of standard costing system would be most appropriate for a manufacturing
business that uses budget targets as part of its employee motivation and reward
package?
4. Compare the relative advantages of ROI and EVA as divisional performance measures.
5. What are the four perspectives of the balanced scorecard as originally recommended
by Kaplan and Norton?
6. Explain how the concept of VFM can be used to measure performance in a public-
sector organization.
7. Explain how performance measures may have unintended behavioural consequences.
Answers on pages 242–243

Exercises
Answers on pages 244–246

Exercise 6.1: Evaluating cost variances


Walkon Co manufactures wooden flooring. The company buys timber which it cuts to standard-length boards,
sands and polishes to sell on to builders. The variance analysis shown in Table 6.6 has been produced for the
production department for the last accounting period.

Table 6.6

$
Material price variance 20,000 (F)
Material usage variance 25,000 (A)
Labour rate of variance 14,000 (A)
Labour efficiency variance 18,000 (F)
Variable overhead expenditure variance 13,000 (A)
Variable overhead efficiency variance 8,000 (F)
Fixed overhead expenditure variance 10,000 (F)

F = favourable variance; A = adverse variance

In response to the variance analysis the production manager has made the following comments:

1. We were experiencing poor staff morale and a high staff turnover so I increased wage rates during the
period. I believe that this has improved staff morale and produced a positive benefit to the company.
2. I was able to source an alternative supplier of raw materials. I negotiated a very good price which I
believe has saved the company a considerable amount of money.
3. We had a large sanding machine which I felt was not being sufficiently used and was therefore costing
the business too much money. I sold this machine and hired a sander only when we needed one.

Required:
Comment on the performance of the production department based upon the variance analysis and the
comments from the production manager provided above.

Exercise 6.2: Total quality management


Coat Co manufactures men’s clothing. The company has for many years operated a traditional standard
costing and variance analysis approach to performance management. The marketing director has recently
been on a training course about quality improvement and has suggested that the company moves to a TQM
approach to performance management. His recommendation will be considered at the next meeting of the
board of directors.

Required:
Write a brief report which considers some of the practical issues that Coat Co would face should the company
decide to change from a standard costing to a TQM approach to performance management.

Exercise 6.3: Non-financial performance indicators


Jake Designs is a small firm which specializes as a consultant in product packaging and marketing within the
cosmetics industry. The company recently appointed a new finance director, Katie Williams. In her first meeting
with the CEO of Jake Designs, Jake McLeod, Katie expressed concerns at the limited focus of the current
performance management system. Katie explained that, although the current system provided good details of
the financial performance of the business, it is also important to include non-financial performance indicators,
particularly those which will provide a better indication of the future performance of the business. She has
suggested that the following performance measures should be reported to the board of directors:

number of customers;
average fees per customer;
average job completion time;
employee turnover rate;
employee job satisfaction;
level of customer satisfaction;
percentage of revenue from new customers.

Jake McLeod is sceptical about Katie’s suggestion. He is concerned that these additional performance
measures will just cause more work for the accounting department and may act as a distraction from more
important tasks such as ensuring that invoices go out promptly and customers pay on time.

Required:
Explain why the inclusion within the performance management system of the non-financial information
suggested by Katie will provide a better indication of potential future success of the business.

Exercise 6.4: Negative behavioural consequences


Tools4U Co operates a chain of tool hire stores across the country. Tool purchases and store staffing levels are
managed centrally, so each store is treated as a revenue centre, with the store manager being responsible for
the level of sales revenue earned, but not the costs. Store managers earn a bonus of 10 per cent of their salary
if the outlet exceeds sales revenue targets for the year.

Required:
Identify how a store manager may be able to manipulate results in order to gain more frequent bonuses.

Answers to comprehension questions


1. The level of budgetary responsibility delegated to an individual manager will depend upon a number of
factors such as organizational structure and management philosophy. Budgetary responsibility is usually
categorized into one of the following four types:
(a) Cost centre – the manager of a cost centre is responsible for managing costs only.
(b) Revenue centre – the manager of a revenue centre is responsible for managing revenues only.
(c) Profit centre – the manager of a profit centre is responsible for managing both revenues and costs,
and therefore the profit of the centre.
(d) Investment centre – the manager of an investment centre will be responsible for costs and revenues
and also the level of investment in non-current assets.
2. A flexible budgeting approach enables the business to adjust its original budget to reflect the actual level
of activity. This approach is useful for businesses with uncertain or unstable levels of activity as it enables
them to fine-tune budgets. This in turn provides more useful levels of information for management.
However, a flexible budgeting approach is not appropriate for an organization that has a fixed level of
funding. Such an organization needs to operate within the confines of the original budget. In such cases,
a fixed budgeting approach would be more appropriate.
3. The effectiveness of budgetary targets in improving employee performance will depend on how
achievable employees perceive the target to be. If targets are too low this may actually pull performance
downwards from where it would have been with no targets at all. If targets are too high then employees
may give up and again performance will be reduced from what it could be. It is therefore important that
standards which provide demanding but achievable budget targets for employees are used.
It would be best for this business to use attainable but demanding standards. If ideal standards were
used, actual performance would consistently fall below budget targets and this would have a
demotivating effect on employees rather than act as an incentive. Basic standards are inappropriate in
this situation as they are unlikely to provide suitably demanding targets for employee motivation.
Standards which are attainable by employees based on current working practices should be set.
However, management can make these demanding in order to maximize employee performance.
4. ROI (return on investment) measures the performance of a division by taking its operating income as a
percentage of the assets employed to earn that income. Because the measure is a percentage, this
enables easy comparison between different divisions and different companies of different sizes.
However, the disadvantage of ROI is that the metric discourages managers from investing in new
projects which offer a return less than existing ROI, even if that project represents a good investment for
the business as a whole.
EVA overcomes this disadvantage of ROI by measuring performance in absolute terms. Any
investment which offers a return above that required by the company as a whole will appear favourable
using the EVA measure. Also, because EVA incorporates a charge for capital, it focuses managers’
awareness on the fact that capital has a cost and that both the income statement and the balance sheet
are important when considering performance levels.
5. The four perspectives of the balanced scorecard are financial, customer, internal and learning and growth
(sometimes referred to as innovation). Kaplan and Norton maintained that the business should measure
its performance from all four of these perspectives and not just from a financial perspective.
6. Value for money can be defined as the optimal use of resources to achieve the organization’s intended
outcomes. This is usually measured using the ‘3E’s’ of economy, effectiveness and efficiency. Economy
refers to the organization successfully operating within given resources. Effectiveness refers to how well
the organization achieves its intended outcomes. Efficiency refers to the relationship between inputs and
outputs.
7. When employees know that their performance is being measured in certain areas, they will concentrate
their efforts on ensuring that they meet these performances and will be less inclined to perform well in
other areas that are not being measured. For example, if the customer services department of a business
is being measured on how quickly it answers telephone calls, it may concentrate on answering and
dealing with calls quickly rather than on the quality of the service provided to customers. Such problems
mean that any performance indicators and targets must be chosen very carefully to avoid negative
behavioural consequences.

Answers to exercises

Exercise 6.1: Evaluating cost variances


The variances reported for the last accounting period appear to reflect some of the decisions made by the
production manager. When assessing these variances it is important to consider the whole picture and the
interrelationships between the different variances.

Material variances
It is clear that the production manager has negotiated a better price on purchases. This has saved the
company $20,000 in material purchase costs. However, the adverse material usage variance suggests that this
cheaper supply of wood may have been of an inferior quality. There will inevitably be wastage in converting
timber into floorboards. In this case the usage variance is $25,000 adverse. This represents a considerable
increase in the amount of wastage. Furthermore, the cost of this additional wastage outweighs the saving in
purchase price of the material. In retrospect, the change to a cheaper supplier may not have been a wise
move.

Labour variances
The production manager gave staff a wage increase during the period and the effectiveness can be seen in the
adverse labour rate variance of $14,000. The production manager claims that this increase has improved staff
morale and performance. The labour efficiency variance appears to bear out this claim as this shows an
$18,000 favourable variance. This labour efficiency variance must also be evaluated in conjunction with the
material usage variance. An adverse material usage variance is often associated with an adverse labour
efficiency variance. In this case the labour efficiency variance is favourable despite the adverse material usage
variance. It would therefore appear that the production manager is correct and the wage increase awarded to
the production staff has had a beneficial impact on the company. The adverse labour rate variance is more
than offset by the favourable labour efficiency variance, with a net benefit of $4,000 for the company.

Overhead variances
The production manager sold the sanding machine which he believed was being under-utilized and therefore
costing the company too much money. The impact of this sale can be seen in the favourable fixed overhead
expenditure variance. No longer owning this machine is saving the company $10,000 per accounting period.
However, in place of this machine the manager is now hiring in sanding equipment when needed. These higher
costs will be reflected in the variable overhead expenditure. As the variable overhead expenditure variance is
$13,000 adverse, this would suggest that the decision was not a good one. Even though the sanding machine
may have been idle much of the time, it was still costing the company less than hiring in the machine when
needed.

Exercise 6.2: Total quality management


Standard costing and TQM involve different cultures and therefore a change from one to the other would
require a change in the mindset of all employees. The most significant cultural change would be a move in
emphasis from one of quantity to one of quality. The focus of standard costing is reducing costs. This can be
totally at odds with a TQM culture in which favourable changes may involve increasing costs.
There will also be a change in allocation of responsibilities within the company. A standard costing system
allocates responsibility for variances to individual departments. If a TQM system were adopted, every employee
would be seen as equally responsible in the quality assurance process.
There would also be changes in attitudes towards waste and idle time. A standard costing system makes
allowances for waste and idle time. This is contrary to the TQM philosophy which aims to eliminate both of
these issues.
The TQM approach is based upon a philosophy of continuous improvement which would involve moving
away from the idea of working to fixed standards. Employees would need to adjust to working in an
environment where regular small changes to processes and procedures were the norm.

Exercise 6.3: Non-financial performance indicators


Katie Williams is effectively suggesting a balanced scorecard approach towards performance management.
This approach involves producing a range of performance measures which cover four important perspectives
on business performance:
the financial perspective;
the customer perspective;
the internal perspective;
the learning and growth perspective.

The reason for including performance measures in relation to these additional perspectives (customer, internal,
and learning and growth) is that these are the factors which will lead to good financial performance in the
future. The problem with financial performance indicators is that they generally only give a measure of past
performance. Good financial performance in the past is no guarantee of good financial performance in the
future.
The business needs to ensure that it is continuing to meet the needs of clients so that they do not go
elsewhere. Likewise, the business needs to look at how it is operating in order to ensure that costs remain low
and operational performance is effective. Non-financial measures are often regarded as indicators of future
financial performance because good performance in these measures will lead to good financial results. For
example, good levels of customer satisfaction will lead to customer retention and the attraction of new
customers. Similarly, high staff morale and low staff turnover result in skill retention and reduced recruitment
and training costs.

Exercise 6.4: Negative behavioural consequences


A store manager may attempt to improve his bonuses by manipulating profit. If the manager anticipated that
sales were going to fall below targets, he might take action in order to boost the level of sales in the year.
As the manager is responsible for sales revenue and not costs, he may be tempted to undertake activities
which boost sales but not necessarily profit. This might include offering discounts to customers in order to
attract more sales, or offering ‘free’ extras such as extended hire or the hire of additional equipment at no
further cost.
If the manager suspected that he was not going to meet his targets as he got close to the year-end, he might
raise some invoices early for orders he knew were going to come in immediately after the year-end. The
manager could raise these invoices before the year-end (and not actually send them out to the customer) so
that the sales will be recorded before the year-end and overall sales for the year increased. This, of course,
represents fraudulent activity and Tools4U Co should implement suitable internal controls to prevent such
practices.
07
Cash flow

OBJECTIVE
To provide an understanding of the importance of cash management to a business and the impact of management strategies
on cash flow and liquidity.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Discuss the importance of cash management to a business.


Interpret a cash budget and identify potential problems.
Formulate strategies for improving cash flow and liquidity.

KEY TOPICS COVERED


The cash-flow cycle.
Information which a cash budget provides, and why this is important.
The format and construction of a cash-flow forecast.
The evaluation of different decisions which impact on cash requirements (financing of asset purchases, credit terms
to customers).
Strategies for improving cash flow.

MANAGEMENT ISSUES
Managers need the ability to evaluate the impact of managerial decisions and strategies on cash requirements. They also
need to be able to analyse cash budgets and use them to make appropriate managerial decisions.

Introduction
Planning and managing cash flows lies at the centre of business success. We hear many
aphorisms which confirm how important this is: ‘cash is king’ and ‘cash is the lifeblood of the
business’ are two commonly used expressions. Why is this the case? Quite simply, a business
cannot survive without cash. It is possible for a business to function for several years without
making a profit provided that it still has cash. But once the cash runs out the business will very
quickly fail. This was the fate of many so-called ‘dot-com’ businesses in the 1990s.

Extract 7.1: Cash flow and small business failure


Dun and Bradstreet, one of the world’s leading credit rating agencies, report that 90 per cent of small business
failures are caused by poor cash flow.

Why does a business need cash?


Cash is needed in a business for a number of reasons: to purchase inventory and to pay wages,
rent, utility bills etc. A business also needs cash to purchase assets such as machinery,
equipment, vehicles and premises. In addition, cash is needed to pay dividends to investors, pay
interest on loans and pay any taxes due. On top of this, at any one time a business may have cash
tied up in current assets such as inventory and accounts receivable. This will be a normal part of
operating the business, but the more money that is tied up in assets the greater the cash
requirement for the business. For example, if a business offers credit to its customers, this will
delay the receipt of cash from sales and will in turn increase the amount of cash the business
needs because it still needs to pay wages and other bills. In a similar fashion, if a business
increases its levels of inventory it will also increase its need for cash.

What is cash flow?


Cash flow refers to the ways in which cash moves in and out of a business through receipts and
payments and also how it circulates within the business to be tied up in various assets such as
inventory of accounts receivable. This flow of cash is represented in Figure 7.1. How quickly
that cash flows and how much cash is tied up in assets will dictate the cash requirements of the
business.
Cash flow within a business is usually categorized into three different aspects:

Operating cash flow: cash flowing in and out of the business from normal day-to-day
operations such as receipts from sales and payments for wages, purchases of inventory,
utility bills and rent.
Investment cash flow: the flow of cash in and out of the business relating to the
purchase and sale of non-current assets.
Financing cash flow: cash inflows from new financing such as new equity or loans and
cash outflows from repayment of financing and payment of interest and dividends.

These three aspects of cash flow are all included within Figure 7.1.

Figure 7.1 Cash flows in a typical business

The cash-flow cycle (sometimes known as the cash conversion cycle) refers to the
circular flow of cash shown at the bottom of Figure 7.1. During the normal operations of a
business, cash gets tied up in working capital: cash is used to pay for purchases and becomes tied
up in inventory. This inventory must be sold and the sales revenue collected from customers
before it is converted back into cash. The cash-flow cycle measures the length of time in days
that it takes for a business to convert purchases into cash. This is an important metric as it can
provide the foundation for establishing how much cash the business needs. The faster cash flows
around this cycle, the less is tied up in working capital and therefore the less cash the business
needs. Figure 7.2 shows this cycle for a typical business.

Figure 7.2 The cash conversion cycle

Figure 7.2 shows a timeline of what typically happens within the business from ordering goods
in to eventually receiving payment for the sale of those goods to a customer. The goods are
ordered in on day 1 but they are not received until day 15. This time lapse between ordering the
goods (day 1) and receiving the goods (day 15) is known as lead time.
If the business has bought the goods on credit, there will be another time lag between
receiving the goods and paying for the supply of them. In Figure 7.2 there is a time lag of 30
days (45 – 15) which is a typical credit period for a supplier. At this point the clock starts ticking
on the cash conversion cycle. The business has paid out money for goods which will not be
converted back into cash until they are sold and the customer has paid. The cash conversion
cycle is therefore the time lapse between paying a supplier and receiving payment from the
customer. In Figure 7.2 the supplier is paid on day 45 and the customer pays on day 227. The
cash conversion cycle is therefore 182 days (227 – 45).

How much cash does a business need?


The amount of cash which a business chooses to hold will always be a trade-off between costs
and benefits. Both holding too much cash and holding too little cash can have a cost to the
business. On the one hand, there is a cost to holding cash and so the more cash the business holds
the greater the cost. This is usually measured as an opportunity cost, that is to say, the return
that could have been earned by investing the cash in other assets. On the other hand, there is also
a cost to raising cash if it is not available when needed. So if a business holds too little cash, this
will also create costs. This will be either the cost of converting assets into cash (by selling them)
or the cost of arranging loans.
Cash, unless it is held in a high-interest investment account, represents shareholders’ capital
that is not being put to work earning a return for the investors. There will therefore be a pressure
to reduce the amount of cash the business holds. The lower the cash balances, the lower the
amount of capital the business uses to operate. Reducing cash will improve the return on capital
employed (ROCE), as demonstrated in Worked Example 7.1.

WORKED EXAMPLE 7.1 Cash balance and ROCE


Quint Co earns a profit before interest and tax of $280,000. The company has total capital of $1,850,000. The current
ROCE is therefore:

Included within the capital of $1,850,000 is a cash balance of $300,000. The finance director decides that the
company does not need to hold this cash and therefore uses it to repay a loan.
This action will reduce the capital of the business to $1,550,000. This in turn increases the ROCE:
The company is therefore able to increase its financial performance (as indicated by ROCE) without any change in
the level of profit simply by reducing its cash balance.

Extract 7.2: Cash mountain at Microsoft


Microsoft, the US software giant, generates over US $20bn in excess cash every year and consequently is
sitting on a mountain of cash. You might think that this is an enviable position to be in, but in fact it is a major
headache for the company. This is an unutilized asset that is substantially lowering the ROCE. How does
Microsoft deal with this? Large cash balances require large investments to use them up and the most
expensive investments are usually acquisitions – buying up other companies. This is why we see Microsoft
making a series of high-cost takeovers – Hotmail, Skype, Firefly, CompareNet, Yammer… the list is long.
Microsoft has recorded over 150 corporate acquisitions since 1987. They need to in order to use up all that
cash.

Although Worked Example 7.1 demonstrates that holding too much cash can have a negative
impact on the business’s financial performance, if the business does not hold sufficient cash it
can experience trading problems (such as overtrading) which will also have a negative impact
on profitability and shareholder returns. This means that a business is constantly seeking to hold
just enough cash to meet its needs, but no more than is necessary. Managing cash is a constant
balancing act between having too much cash and not enough.

Expert view 7.1: Cash flow and overtrading


Overtrading is a problem which hits many businesses. It occurs when the business tries to grow too quickly and
has insufficient funds to pay the increasing costs which accompany an increase in sales. A growing order book
may seem like good news, but it also creates a demand for more cash within the business. Meeting growing
levels of sales also requires increases in working capital: more inventory and more money tied up in trade
receivables. Extra orders may also mean overtime payments to employees and the purchase of more
equipment. A business therefore needs to manage its growth carefully to ensure that it can be funded.

The amount of cash which an individual business needs to hold will depend upon the nature of
its operations and the attitudes of its managers. Financial theory categorizes the reasons for
holding cash into four ‘motives’:

the transaction motive;


the precautionary motive;
the speculative motive; and
the compensating balances motive.

The transaction motive


Any business will need to hold a certain amount of cash in order to pay the day-to-day bills. For
most businesses, there will be a gap between receiving cash from sales and needing cash to make
payments to employees and for purchases, utilities and equipment. The greater this gap, the more
cash the business will need to hold. This cash may be held directly by the business (in cash
registers and secure safes) or may be held in non-interest-bearing bank accounts. Some
businesses hold a balance of cash in deposit accounts which earns some interest. Because such
accounts provide easy access to cash, interest rates are usually low; better returns will be
available elsewhere.

The precautionary motive


Most businesses will wish to hold a certain reserve of cash in case of emergencies. This will
include unforeseen and unexpected expenses and may be referred to as ‘financial slack’. If a
business does not have a certain amount of slack in its cash balances, it may have to raise money
quickly and in the short term if an unexpected need arises. This could prove to be very costly.
Research has shown that although holding precautionary cash balances has a cost, this is less
than the cost of having to raise cash at short notice when an emergency arises.

Extract 7.3: Cash holdings and the rise of the precautionary motive
You might expect that the improvements in information and financial systems over recent years have enabled
businesses to reduce their cash holdings. In fact, the opposite is true. The average cash to assets ratio of
industrial companies in the United States has more than doubled since 1980. Research has identified that this
trend correlates with industry risk – ie businesses are holding more cash to protect themselves against riskier
cash flows: the precautionary motive. The situation is amplified by the fact that new manufacturing technologies
have reduced inventories and receivables, thus tying up less cash in other assets.

The speculative motive


Businesses will also hold cash in order to be able to take advantage of unexpected opportunities
such as new investments, changes in interest rates and favourable fluctuations in exchange rates.
Good investments may be missed if the business does not have funding readily available. If
interest rates are low, businesses will typically maintain higher speculative balances of cash; if
interest rates are high, businesses will commit these balances to high-yield investments.

Extract 7.4: ING Direct and Emirates – speculative gains


Businesses that have sufficient cash funds available at the right time are able to benefit from unexpected
opportunities. The benefit is often greater if competitors are not in a similar position to take up the opportunity:

At the time of the Icelandic banking crisis, ING Direct snapped up the deposits being offloaded by failing
Icelandic banks.
Emirates bought up the new Airbus A380 at a time when other airlines had insufficient funds due to the
industry downturn after 11 September 2001. This gave Emirates a competitive advantage as the A380
had greater range, passenger capacity and fuel economy than the planes being used by competitors.

The compensating balances motive


Most businesses that maintain current accounts with commercial banks are required to maintain a
minimum balance in their account. This minimum balance, referred to as the compensating
balance, is to compensate the bank for the services they provide and essentially provides the
bank with free use of the business’s money. If such a balance is not maintained, higher fees are
normally imposed by the bank.

Methods of establishing cash balances


A number of models have been developed to assist managers in establishing appropriate cash
levels. In this chapter we will look at the two most popular and widely used models: the
Baumol–Tobin model and the Miller–Orr model. Each model takes a very different approach to
establishing the cash balance that a business should hold.

The Baumol–Tobin model of cash management


The Baumol–Tobin model is a transaction-based model which assumes that a business’s demand
for cash is consistent over time and can be predicted with certainty. As a business uses its
available cash it will need to convert investments into cash by selling them. This will incur
transaction costs and therefore there is an incentive to minimize the number of such sales.
However, if large quantities of cash are converted each time, there will be a high opportunity
cost of lost interest because of holding cash balances rather than investments. The model
therefore attempts to establish optimum cash balances by determining the amount of cash that
should be converted from investments each time cash is required.
The Baumol–Tobin model therefore attempts to establish optimum cash balances by
quantifying the trade-off between the opportunity costs of holding too much cash and the
transaction costs involved in holding too little cash. This trade-off can be represented
diagrammatically as shown in Figure 7.3.

Figure 7.3 The cost of holding cash


The opportunity costs represent the interest forgone by holding funds as cash rather than
placing them in an investment. These costs rise with the size of the cash balance.
The trading costs represent the costs of having to raise cash in the short term either by
liquidating assets or by taking out a loan. The lower the size of the cash balance, the higher these
trading costs will be.
If frequent small conversions to cash are made, average cash balances will be low and
opportunity costs minimized, but trading costs will be high owing to the high frequency of
transactions. See Figure 7.4.

Figure 7.4 Frequent cash conversions

On the other hand, if large quantities of investments are converted to cash, trading costs will
be minimized but average cash balances and therefore opportunity costs will be high. See Figure
7.5.

Figure 7.5 Infrequent cash conversions


Using the Baumol–Tobin model, the business attempts to minimize the total costs of holding
cash, which will be made up of both trading costs and opportunity costs. Thus, the optimal cash
balance (ie that which has the lowest total cost to the business) is found where the opportunity
cost equals the trading cost (see Figure 7.3).
The Baumol–Tobin model determines the optimum quantity of cash to convert each time cash
is required. This quantity is represented mathematically by the expression:

WORKED EXAMPLE 7.2 The Baumol–Tobin model


The following information relates to Ergin Co: the fixed cost of selling assets to raise cash is $500; cash outflows
exceed cash inflows by $9,000 per week; the interest rate earned on marketable securities is 5 per cent.
What is the optimum cash balance for Ergin Co?

Answer
Using the Baumol–Tobin model:

So when the Ergin Co requires cash, it should convert $13,416. This means that the average cash balance carried by
the business will be approximately $6,708 (13,416 ÷ 2).

The Baumol–Tobin model suggests that when interest rates are high, cash balances should be
kept to a minimum. However, the model is based upon a number of assumptions, which reduces
its usefulness in complex real-world situations. The model assumes that:
Cash outflows are predictable and even over time.
Cash inflows are predictable and regular.
Day-to-day cash needs are funded from a bank current account.
A ‘buffer’ of short-term investments is held which can be liquidated to provide cash when
needed.

In practice these assumptions do not hold true for most businesses. In particular, cash inflows
and outflows are not even and regular and can be difficult to predict. Therefore a more useful
model for cash management is one which allows for irregularity and fluctuations in cash flows.
One such model is the Miller–Orr model.

The Miller–Orr model of cash management


The Miller–Orr model provides a method of cash management for businesses that do not have
uniform cash flow and therefore find it difficult to predict levels and timing of cash inflows and
outflows. The model allows for daily variations in the cash balance within prescribed control
limits known as the upper limit and the lower limit.
The model is applied by firstly deciding upon a minimum acceptable cash balance (the lower
limit). This is set by management and will be used to establish a buffer of cash for emergencies.
The lower limit will depend upon how much risk of a cash shortfall management are willing to
accept. This in turn will depend upon the consequences (and costs) of experiencing a cash
shortfall together with how quickly and easily the business can access additional funds.
Having established this minimum cash balance, the Miller–Orr model can be used to establish
a maximum cash balance (the upper limit) and a return point. The return point represents the
optimum cash balance. Cash is then managed as follows:

If cash levels reach the upper limit, an amount is transferred into investments in order to bring
the cash balance back down to the return point.
If cash levels drop to the minimum level, an amount is converted from investments into cash
to bring the cash balance back up to the return point.

This principle is demonstrated in the diagram in Figure 7.6.

Figure 7.6 Cash limits


The difference between the upper limit and the lower limit is known as the spread. This is
calculated using the formula shown in Figure 7.7.

Figure 7.7 The spread formula

In the figure:

The transaction costs represent the cost of converting short-term investments into cash or
vice versa.
The variance of cash flows is a measure of the fluctuation in the cash flows and therefore
cash balances. The assumption underlying the Miller–Orr model is that although cash
balances can fluctuate randomly, these fluctuations are normally distributed. It is therefore
possible to calculate a mean balance and standard deviation of cash balances based upon
past data (Variance = Standard deviation2).
The interest rate represents the return that can be earned by converting cash into marketable
securities or other similar investments.

Both the variance and interest rates should be expressed in daily terms.
Once the spread is calculated, the return point can be found using the following formula:

Return point = Lower limit + (1/3 × spread)

We will now see how this model is applied using a simple illustration in Worked Example 7.3.

WORKED EXAMPLE 7.3 The Miller–Orr model


Zetoc Co sets its minimum cash balance at $5,000 and estimates the following:

Transaction costs = $500 per sale or purchase of short-term investment.


Standard deviation of cash flows = $1,200 per day (therefore variance = $1.44m per day).
Interest rate = 7.3 per cent per year ( = 0.02 per cent per day).

Given the lower limit of $5,000, what upper limit and return point should Zetoc Co set for its cash management?

Answer

Using the formulae above:

Spread = 3[(3/4 × $500 × $1.44m) ÷ 0.02%]1/3 = $41,774


Therefore the upper limit = $5,000 + $41,774 = $46,774
And the return point = $5,000 + (1/3 × $41,774) = $18,925

This means that Zetoc Co should try to maintain its cash balance at around $18,925 (in practice this will probably be
rounded to $19,000 or even $20,000 for convenience).
If the cash balance rises as high as $46,774, the company should transfer $27,849 ($46,774 – $18,925) from the
bank account into short-term investments in order to bring the balance back down to $18,925. On the other hand, if
the cash balance falls as low as $5,000, the company should convert $13,925 ($18,925 – $5,000) of short-term
investments back into cash to bring the balance up to $18,925.

Exercises: now attempt Exercise 7.1 on page 285

Cash forecasting: the cash budget


A business needs to produce a cash budget alongside the budgeted income statement and balance
sheet in order to ensure that the business holds sufficient cash to achieve the targets set within
the operational budget. A cash budget (sometimes referred to as a cash-flow forecast) shows
the amount and timing of cash receipts and cash payments. In doing so, it also predicts the
amount of cash the business will need.
A cash budget may be produced on a daily, weekly or monthly basis dependent upon the
volatility of cash flows of the business and the need for level of detail. In this section we will
show you how the cash budget is a useful tool for planning the business’s future cash needs and
how it can be used to inform management strategies.

The format and construction of a cash budget


Cash-flow planning is central to good cash management. It is therefore important that managers
are able to produce and understand cash budgets. The following example demonstrates how to
produce a simple cash budget for a business.
WORKED EXAMPLE 7.4 A simple cash budget
Pascal starts a furniture-making business. Budgeted sales and purchases for the first six months of his business are:

Jan Feb Mar Apr May Jun


$ $ $ $ $ $
Sales 5,000 6,000 6,500 7,000 6,000 7,000
Purchases 4,000 6,000 4,000 4,000 3,500 4,000

1. Pascal anticipates that 50 per cent of his sales will be for cash and 50 per cent on credit terms of one month.
2. Inventory will be kept very low and bought in the month it is sold. Inventory is paid for with cash as it is bought.
3. Workshop rent is $1,200 per month, paid on the first day of each month.
4. Pascal has a van. This costs $400 per month. Pascal pays all van costs in cash.
5. Pascal starts the business with an opening bank balance of $5,000.

Required:
Prepare a cash budget for the first six months of Pascal’s business.

Answer:
There are a number of logical steps in producing a cash budget:

Forecast sales.
Forecast cash receipts from sales.
Forecast cash payments.
Estimate end-of-period cash balances.

In this simple example we will show you how to carry out each of the steps and to compile the information into a cash
budget.

Forecast sales
When preparing a cash budget it is necessary to determine cash receipts in two steps. Firstly, a budget must be
established for sales. In the second step this sales forecast is translated into a cash receipts forecast. The reason for
these two steps is that the cash budget is concerned with the timing of cash receipts and sales do not necessarily
translate directly into cash, particularly if sales are made on credit.
Worked Example 7.4 is relatively straightforward and we are given predicted sales figures. In practice, these would
have to be established based upon market research, known contracts or past trends. A variety of statistical and
mathematical techniques may be used in forecasting sales.

Jan Feb Mar Apr May Jun


$ $ $ $ $ $
Sales 5,000 6,000 6,500 7,000 6,000 7,000

For many businesses the pattern of sales is not even throughout the year, but rather is seasonal. For example, 40 per
cent of toy sales occur in the six weeks before Christmas. Accurately establishing this pattern can be extremely
important for cash-flow planning.

Forecast cash receipts from sales


Once a sales forecast has been prepared, it is necessary to determine when those sales will be collected as cash
receipts. If sales are made on credit, there will be a delay between the point of sale and the point of cash receipt.
In Worked Example 7.4 Pascal predicts that 50 per cent of his sales will be for cash and 50 per cent on one
month’s credit. Based upon this prediction we can establish when cash from sales will be received. For the purposes
of this exercise we will assume that cash from cash sales is received immediately and that cash from credit sales is
received in the following month. In practice, all customers may not pay on time and some cash receipts may be even
later. However, we will keep it simple for this example.
In the table below, the forecast sales for each month are split so that 50 per cent of the money is received in that
month (the cash sales) and 50 per cent is received in the following month (the credit sales). For example, January’s
forecast sales of $5,000 will be received in cash as $2,500 in January and $2,500 in February.

Jan Feb Mar Apr May Jun


$ $ $ $ $ $
Sales 5,000 6,000 6,500 7,000 6,000 7,000
Cash sales 2,500 3,000 3,250 3,500 3,000 3,500
Credit sales 0 2,500 3,000 3,250 3,500 3,000
Total receipts 2,500 5,500 6,250 6,750 6,500 6,500

Of course, 50 per cent of the cash from June’s sales will be received in July, but as this lies outside our six-month
budget period we can ignore this.

Forecast cash payments


Cash payments should be recorded when they are expected to be made. You should start with those payments that
are fixed amounts due on known dates, such as loan repayments, interest charges, rent or any other standing
payments. Payments to creditors should be scheduled based upon predicted purchase patterns and known credit
periods.
In Worked Example 7.4 we are given a schedule of purchases and told that these purchases are made in cash. We
are also given information about workshop rent and van running costs. This information can be used to produce a
schedule of cash payments as set out below:

Payments: Jan Feb Mar Apr May Jun


$ $ $ $ $ $
Inventory 4,000 6,000 4,000 4,000 3,500 4,000
Rent 1,200 1,200 1,200 1,200 1,200 1,200
Van 400 400 400 400 400 400
Total 5,600 7,600 5,600 5,600 5,100 5,600

Estimate end-of-period cash balances


Finally, we can combine our cash receipts and cash payments forecasts to establish the net increase or decrease in
cash balance each month. This allows us to estimate end-of-period cash balances. In Worked Example 7.4 we are
told that the opening cash balance will be $5,000. Using this information we can now create a full cash budget. This is
set out in Table 7.1.

Table 7.1 Cash budget for Pascal

Jan Feb Mar Apr May Jun


$ $ $ $ $ $
Cash receipts
Cash sales 2,500 3,000 3,250 3,500 3,000 3,500
Credit sales 0 2,500 3,000 3,250 3,500 3,000
Total receipts 2,500 5,500 6,250 6,750 6,500 6,500
Cash payments
Inventory 4,000 6,000 4,000 4,000 3,500 4,000
Rent 1,200 1,200 1,200 1,200 1,200 1,200
Van 400 400 400 400 400 400
Total payments 5,600 7,600 5,600 5,600 5,100 5,600
Increase/(Decrease) (3,100) (2,100) 650 1,150 1,400 900
Opening balance 5,000 1,900 (200) 450 1,600 3,000
Closing balance 1,900 (200) 450 1,600 3,000 3,900

Table 7.1 shows the total expected cash receipts and cash payments in each month and the resultant increase or
decrease in the cash balance. For example, in January total receipts are $2,500 whereas total payments are $5,600.
This will result in a decrease in cash balance of $3,100. As the opening cash balance is $5,000, this means that the
closing cash balance will be only $1,900. As $1,900 is the closing cash balance in January, it will also be the opening
cash balance for February.
It can be seen from this cash budget that if the projected pattern of cash receipts and payments occurs, Pascal will
run out of money by the end of February when the closing cash balance is predicted to be negative $200. This means
that either he will have to put in place arrangements to borrow money in February or he will have to change his plans
somehow to avoid this negative cash situation. However, we can see from the forecast that this situation is only
temporary and by the end of March cash balances should be positive again until the end of the six-month period of
the forecast.
It is the fact that the cash budget allows us to identify potential problems such as this in advance that makes them
so useful. In the next section we will look in more detail at how cash management can be improved by concentrating
on the key areas that cause cash problems for businesses.

Exercises: now attempt Exercise 7.2 on page 285

Cash management: strategies for improving cash flow


There are certain key areas that always have the potential to cause cash problems for any
organization. Good cash management therefore needs to focus on these four key areas:

managing accounts receivable;


managing accounts payable;
managing inventory;
strategic financing of assets.

The first three areas in this list are concerned with the amount of money tied up in working
capital. Managing these three areas is therefore known as working capital management.
Good working capital management can substantially reduce the amount of cash a business needs.
The fourth area of asset financing is equally important. Research has shown that one of the
greatest growth barriers to businesses is an inability to finance new assets for growth. We will
therefore examine different methods of financing assets and how these impact upon cash flows.

Managing accounts receivable


Extract 7.5: The reality of offering credit
Around 90 per cent of businesses that operate in industry or wholesale offer credit to their customers. In
addition, around 40 per cent of retail sales are made on credit. This means that accounts receivable are a cash
management issue for the vast majority of businesses.

Good management of accounts receivable can substantially reduce the cash needs of a business.
On the other hand, poor management of accounts receivable is often a source of cash-flow
problems, particularly in small businesses. Effective management of accounts receivable
involves four key steps:

formulating an appropriate credit policy;


setting credit levels;
credit control; and
debt collection.

Credit policy
Cash flow can be improved by offering shorter credit periods to customers and thereby reducing
the amount of cash tied up in accounts receivable. The best credit policy from a cash-flow point
of view is to offer no credit and trade in cash sales only. That would mean that cash is received
as soon as a sale is made. However, in many industries it is common practice to offer credit to
customers and a business would not be able to trade if it did not offer credit terms similar to
those of competitors.
Many businesses offer early settlement discounts on their invoices to encourage customers to
pay earlier. Although this technique can be successful, it is not necessarily financially beneficial
to the business. The discount offered must be weighed up against the cost of the outstanding
debt, as illustrated in Worked Example 7.5.

WORKED EXAMPLE 7.5 Credit policy and early settlement discounts


Mahmoud Co trades in an industry in which it is usual practice to offer 30 days’ credit on all sales. Analysis of
Mahmoud Co’s sales ledger reveals that customers on average actually take 38 days to pay.
The sales director has suggested that the company increase its credit period to 60 days, which he says will make
Mahmoud Co more attractive to customers and will increase sales by 5 per cent. The finance director has suggested
that if the credit period is extended, the company should also offer an early settlement discount of 2 per cent for
payments made within 14 days. He believes that 25 per cent of customers will take advantage of the early settlement
discounts and that this will counterbalance the extended credit taken by other customers.
The company has annual sales revenue of $5 million and earns a contribution of 40 per cent on sales. Accounts
receivable are financed by an overdraft which has an annual interest rate of 6 per cent.
Should the company change its credit policy in the way suggested by the sales director and the finance director?

Answer
The current collection period on accounts receivable is 38 days. This means that the average balance on accounts
receivable is $520,548 ($5m sales × 38/365).
If the new credit policy is implemented, the average collection period will increase to (25% × 14 days) + (75% × 60
days) = 49 days. Sales will also increase by 5 cent to $5.25 million per year. The level of accounts receivable will
therefore increase to: $5.25m × 49/365 = $704,795.
With this increase in outstanding accounts receivable, financing costs will increase by ($704,795 – $520,548) × 6%
= $11,055 per annum. The discount offered to early settlers will cost $5.25m × 25% × 2% = $26,250. However, from
the 5 per cent increase in sales the company will earn additional contribution of $250,000 × 40% = $100,000.
Therefore, there will be a net financial benefit to the company of $62,695 ($100,000 – $26,250 – $11,055) by
changing the credit policy and offering early settlement discount. Mahmoud Co should implement the policy change.

Credit analysis
The second step in good credit management is to set appropriate credit levels for customers. This
involves assessing the creditworthiness of customers and deciding how much credit you are
going to extend to them.
Businesses should make use of credit references and other sources of information. They
should analyse the accounts of potential new customers for signs of liquidity problems and
potential bankruptcy problems.
Assessing creditworthiness is not just a one-off task for new customers. Credit level should be
constantly reviewed and checks frequently made on customers. Sales staff can be very helpful in
obtaining information about customers, such as rumours and reputation within the industry. They
are also in regular contact with customers and therefore are able to feed back their impressions
from dealings with the customers and visits to their premises.
It is frequent practice to extend little or no credit to new customers and to gradually extend
credit limits as customers prove their creditworthiness by prompt payment of invoices.

Credit control
Once credit limits have been set, they need to be constantly monitored and controlled. There are
two aspects to this task. Firstly, the company must ensure that no customer exceeds their credit
limit. Before any new order is accepted the current outstanding balance from the customer
should be checked to make sure that the new order does not take the customer over their credit
limit. For example, if a customer had a credit limit of $15,000 and currently had outstanding
invoices of $12,000, an order worth $4,000 would not be processed until the customer had paid
off at least $1,000 of the outstanding balance.
The second aspect of good credit control is ensuring that customers pay on time. One of the
most frequently used tools for this task is the aged debtors report. This is a report generated
from the sales ledger which shows the outstanding balance from each customer broken down by
age of the debt. An extract from an aged debtors report is shown in Table 7.2.

Table 7.2 Aged debtors report

Customer Total <30 days 30–60 days 60–90 days >90 days
$ $ $ $ $
Adam Co 15,981.81 15,981.81 – – –
Benson Co 27,613.59 – – 18,279.03 9,334.56
Cantor Co 18,516.37 17,935.70 580.67 –

Table 7.2 shows an extract from an aged debtors report for a company which offers 30 days’
credit to customers. This report enables managers to identify if there are problems with any
customers and to initiate action to chase outstanding debts. In the report it can be seen that the
first account, Adam Co, has no amounts outstanding outside of the 30-day credit period.
However, Benson Co and Cantor Co both have outstanding invoices which have not been paid
within the credit period.
Benson Co has an outstanding balance of $27,613.59, all of which is more than 60 days old.
This would suggest that there is a problem with this customer. No further credit sales should be
made to the customer and action should be taken to recover the outstanding balance.
Cantor Co appears to be paying invoices within the credit period, with the exception of one
small invoice for $580.67. This might suggest that there is a dispute on the invoice. This needs to
be investigated and resolved.

Debt collection
Even with a good credit policy in place and frequent credit checks on customers, it is necessary
to ensure that debts are collected promptly. The longer a debt is outstanding, the greater the
likelihood that the customer will never pay and the business will incur a bad debt.
To ensure prompt payment, invoices should be sent promptly. Ideally, invoices should be sent
as soon as goods are shipped. If an account becomes overdue an action should be taken
immediately. There is a series of escalating follow-up actions which can be taken if an invoice is
not paid on time:

Send the customer a statement of their account which highlights overdue balances and late
payment penalties.
Telephone the customer to ask for payment.
Send a formal letter to the customer reminding them that payment is overdue and late
payment penalties will be incurred.
Visit the customer in person (this will usually be done by a member of the sales department)
to discuss the outstanding balance and secure payment.
Send a solicitor’s letter threatening legal action if the balance is not paid.
Hand the debt over to a debt collection agency that will enforce payment on behalf of the
company.

If the customer is consistently failing to pay invoices on time, the business should restrict credit
to that customer until all outstanding bills are cleared.

Invoice discounting and debt factoring


Even if a business follows all of the steps above for good credit management it may still find
itself having a large amount of money tied up in accounts receivable. For example, if the
business operates in an industry in which it is necessary to offer credit to customers, it may not
be able to avoid having large balances of accounts receivable. If this is the case, there are some
ways in which the business can free up this cash.
One such method is invoice discounting. This involves using accounts receivable as
collateral against borrowing from a finance company. The finance company will typically lend
up to 80 per cent of the value of outstanding sales invoices. As invoices are paid and new ones
are issued, the amount of borrowing available will be adjusted to maintain that fixed percentage.
Interest will be charged on the amount borrowed and usually the finance company also charges a
monthly service fee.
In order for invoice discounting to work, the finance company must be confident of the
creditworthiness of customers. Borrowing available is rarely greater than 80 per cent of total
outstanding invoices as this provides the finance company with a margin for bad debts. Also, a
finance company may refuse to lend against certain invoices if it feels that they represent a high
risk. As security, the finance company providing the loan will take a floating charge over the
accounts receivable of the business.

Expert view 7.2: Fixed and floating charges


Anyone lending money to a company will want some security against that loan. It is therefore usual to take a
charge against an asset of the company. For example, a lender may take a charge against one of the
business’s buildings. In the event of default on the loan the lender has the legal right to seize the asset, sell it
and recoup the amount they are owed. If this charge is against a clearly identified asset, it is known as a fixed
charge. However, in the case of some assets such as accounts receivable it will be difficult to fix the charge
against any one asset as individual invoices are constantly being raised and paid off. In this case the lender will
raise a floating charge over the class of assets in general.

Invoice discounting can prove to be a very effective and flexible means of raising extra cash for a
business that has no other options. The business will only pay interest on the amount it borrows,
which will not necessarily be the full amount available. The system essentially works rather like
an overdraft but it is usually more expensive than a bank loan or overdraft and therefore would
only be used if these were not available.
Another method of releasing cash from the sales ledger is debt factoring. Unlike invoice
discounting, which is borrowing against the value of accounts receivable, debt factoring involves
actually selling sales invoices on to a third party called a factor. Some banks offer factoring
services or the factor may be a specialist financial company. If an invoice is sold to a factor a
cash advance (typically around 80 per cent of the invoice value) is made immediately. The
balance of the invoice value less the factor’s fee is then paid upon settlement of the invoice by
the business’s customer. The factor may or may not take on the risk of bad debts. Usually this is
an optional extra which carries an increased fee.
The debt-factoring company may also take on responsibility for managing the sales ledger and
chasing and collecting unpaid sales invoices. Although there is obviously an additional cost of
this service, it can be beneficial as a form of outsourcing; as the factor has expertise in this task it
can reduce bad debts, speed up debt collection and free management time to concentrate on core
activities.
A related but less frequent method of raising cash against accounts receivable is forfaiting.
This is sometimes used by exporting businesses and involves selling individual export sales
invoices using a method similar to debt factoring.

Exercises: now attempt Exercise 7.3 on page 286

Managing accounts payable


Just as a business needs to collect its accounts receivable as soon as possible in order to improve
its cash flow, it can also improve cash flow by paying creditors as late as possible. The ultimate
cash-flow position for a business is to buy on credit and sell for cash, such that cash is received
from customers before payments must be made to suppliers. In this way, sales finance purchases.
Unfortunately, for most businesses this is not the case and so extra cash is needed to ‘buffer’ the
timing difference between converting accounts receivable to cash and having to pay creditors.
The longer the cash conversion cycle (the time gap between collecting accounts receivable
and paying creditors), the more cash the business will need. Therefore, good management of
accounts payable can involve extending credit periods as far as possible. In the first instance, the
business should seek to get favourable credit terms from its suppliers. It should then carefully
manage its payments to stretch these credit terms as far as practicable. However, the business
should always take care against stretching it too far as it could result in the supplier suspending
the account. This could disrupt trading and also result in poor credit ratings which cause the
business difficulty in setting up accounts with new suppliers.
If suppliers offer discounts for early payment, the business needs to balance the advantage of
reduced purchase costs against the cost of having to make the payment earlier.
Good management of accounts payable also involves ensuring that only valid invoices are
paid. It is therefore important that the business has in place a good system of internal controls
to check physical receipts of inventory against invoices and to note any omissions or damage.

WORKED EXAMPLE 7.6 Should you take a cash discount?


Red Co purchases supplies at a cost of $5,000. The supplier’s standard credit terms are 30 days, but the supplier
offers a 2 per cent discount if the invoice is paid within 10 days. Red Co has an overdraft which incurs an interest
charge of 0.05 per cent per day.
Should Red Co take the early payment discount?
Answer
If Red Co takes the early payment discount it will save $100 on its purchases. However, it will also incur higher
interest charges on the overdraft as follows:

$4,900 × 20 days × 0.05% = $49

As the early payment discount outweighs the additional overdraft interest charges, Red Co should make the early
payment. It will be $51 ($100 – $49) better off by doing so.

Managing inventory
Holding inventory can be very expensive, particularly if the goods held are of high value or large
quantities are involved. This can mean a substantial amount of cash tied up in inventory. So why
do businesses carry inventory? In practice, there are a number of reasons:

Stocks of finished goods may be held as a buffer where there is fluctuation and uncertainty
in patterns of demand. If it is difficult to predict patterns of demand, a high level of
inventory will be necessary to ensure that inventory does not run out at those times when
demand is high. Some businesses with seasonal sales (such as firework manufacturers) will
carry high levels of inventory throughout much of the year because they will be producing
throughout the year for sales over a very short period.
Running out of inventory may carry a cost to the business. If no stock is available customers
will buy from a competitor. This could mean not only a lost sale but also a lost customer.
Some businesses manufacture in batches, such that they will produce a large quantity of one
item and then retool the production system to run a batch of another inventory item. This
method of operating may be necessary to achieve economies of scale through large runs of
production or it may be a requirement of the manufacturing process. In either case, it will
result in the business carrying inventory to ensure that all items are available even when
they are not being produced.
A business may purchase large quantities of raw materials at a time in order to take
advantage of bulk discounts. This will result in the business carrying inventory.
Most retailing businesses need to carry sufficient inventory for the shop to look full in order
to be attractive to customers.

The amount of inventory carried needs to be just right – not too much and not too little, as each
carries a cost to the business.
The costs of carrying too much inventory include:

the opportunity cost of having money tied up in inventory;


higher storage costs;
increased handling and insurance costs;
increased costs of inventory deterioration, obsolescence and theft.

On the other hand, the costs of carrying too little inventory include:
the cost of unfulfilled orders;
the cost of lost customers;
idle machines and employees if raw material inventory runs out.

Inventory management
Good inventory management is a matter of minimizing the costs associated with holding
inventory, whether they be the costs of holding too much or the costs of holding too little
inventory. There are many inventory control models which tell a business how much inventory
to order and when to order it in order to minimize costs. In this chapter we will look at four
approaches to inventory management:

the inventory turnover ratio;


inventory reorder levels;
ABC inventory management;
the economic order quantity model;
just-in-time inventory management.

The inventory turnover ratio


The most common method of analysing and assessing inventory levels is to use the inventory
turnover ratio. This ratio shows a business how many days’ worth of inventory it is holding on
average throughout the year. It is expressed by the following formula:

Inventory can be measured against sales turnover, but it is better to use cost of sales because
inventory is recorded at cost and therefore you are comparing like with like. Also, it is possible
to use the end-of-year inventory balance but if inventory levels change seasonally, using an
average balance is a more accurate measure.
The inventory turnover ratio can be useful for benchmarking the level of inventory the
business holds in two ways:

Firstly, inventory holdings can be analysed over time to see if the inventory levels are
changing in relation to the level of sales (as reflected in cost of sales). If inventory days are
increasing, this may suggest unnecessary stockpiling or it may reveal that the business is
holding obsolete and unsaleable inventory. If inventory levels are falling over time, this
may reflect more efficient inventory management or it could be a result of inefficient
management which may result in stockouts and lost sales.
Secondly, the ratio enables comparison with industry rivals. If the business finds that it is
holding inventory for longer than its competitors, it should be concerned that it is not
operating as efficiently as it could in its inventory management. Too much cash is being tied
up in inventory and this has a cost for the business.
WORKED EXAMPLE 7.7 Comparing inventory turnover
The following data are being extracted from the most recent financial statements of Ant Co and its two main
competitors, Ben Co and Crab Co:

Ant Co Ben Co Crab Co


$m $m $m
Cost of sales 140.0 50.2 300.8
Opening inventory 17.5 4.2 21.8
Closing inventory 19.1 4.6 23.2

Using these data, the respective inventory turnover ratios can be calculated:

This analysis shows that Ant Co is holding its inventory for considerably longer than its competitors. This would
suggest that it is carrying too much inventory and needs to adjust its inventory management policy.

Lead time and buffer stock


If a business uses inventory in a fairly regular and consistent manner, its inventory levels will
look like the graph in Figure 7.8.

Figure 7.8 Inventory reorder level


The business will manage its inventory level in the following manner:

Firstly, it will establish a buffer stock which represents the minimum inventory it wishes
to carry. This is an insurance against stockouts and the disruption which they would cause.
Secondly, the business needs to calculate the lead time on receiving new inventory. This is
the time delay between placing an order and actually receiving delivery of the inventory.
Using this information, the business can work backwards from the buffer stock level, using
the lead time, to establish the level of inventory at which a new order should be placed – the
reorder level. As inventory falls to this level the inventory management system should
trigger an order to the supplier.

Many businesses use a computerized inventory control system which is linked to the ordering
system and in some cases also linked to the supplier’s sales and stock dispatch systems so that
the whole process is automated. This automation can substantially reduce the lead time on
inventory delivery and consequentially reduce the quantity of inventory that the business needs
to carry. When taken to the extreme, this results in just-in-time inventory management (see
below).

Economic order quantity (EOQ) model


The EOQ model is an approach to inventory management which attempts to minimize the costs
of holding inventory. The model recognizes that there are two opposing sets of costs which need
to be balanced and attempts to do so by calculating the optimum quantity of inventory to order at
any one time:

If large quantities of inventory are ordered at a time, the business will incur higher costs in
storage and handling; there will be a higher cost of inventory obsolescence, damage or loss;
and the opportunity cost of having cash tied up in inventory will be greater.
If small quantities of inventory are ordered each time, orders will need to be placed more
frequently and levels monitored more closely. This will increase order costs.

Figure 7.9 illustrates how these opposing sets of costs lead to a total cost of stockholding which
has an optimum (ie minimum) cost when storage costs = order costs.

Figure 7.9 The economic order quantity

This relationship can be expressed with the following formula:

Therefore, when using the EOQ model, the amount of inventory held by the business will be
dictated by the most economic quantity to order. This will be the quantity which minimizes the
total cost of holding inventory. The method is illustrated in Worked Example 7.8.

WORKED EXAMPLE 7.8 Optimizing inventory order levels


The management of Pantar Co wishes to minimize the business’s inventory costs. The company accountant has
identified the following costs and figures in relation to inventory item P31:

Annual demand for P31 is 80,000 units.


P31 costs $3.00 per unit.
Inventory management costs for P31 are:
– Ordering cost: $15.00 per order
– Holding cost: $2 per unit per year.

Required:
Calculate the total cost of inventory for P31 if Pantar Co applies the EOQ.
Answer
The EOQ for P31 will be:

= √(2 × 15 × 80,000/0.5)/2 = 1,095 units

Therefore, the number of orders placed in a year will be:

= 80,000/1,095 = 73 orders per year

The annual ordering cost:

= 73 × $15 = $1,095 per year

The average inventory of P31 held by Pantar Co will be:

= 1,095/2 = 547.5 units

The annual holding cost:

= 547.5 × $2 = $1,095 per year

(Note that the annual ordering cost = annual holding cost at the optimum level as illustrated in Figure 7.9.)

Inventory purchase cost:

= 80,000 × $3 = $240,000

Total cost of inventory with EOQ policy:

= $240,000 + $1,095 + $1,095 = $242,190 per year

ABC inventory management


ABC inventory analysis is a method of prioritizing different inventory lines in order to focus
management attention on the most critical areas of inventory. It employs the Pareto Principle
– the principle that 80 per cent of overall revenue is based upon only 20 per cent of inventory
items. This means that not all inventory lines are equally important. Efficient management means
focusing on the most important lines and less on other areas of inventory. For this reason the
technique is also sometimes known as selective inventory control.
When applying the ABC approach all inventory is evaluated and divided into three categories
– A, B or C – as illustrated in Table 7.3.

Table 7.3 ABC Inventory analysis

Category Inventory quantity (%) Inventory value (%)


A 12 72
B 28 18
C 60 10
Total 100 100

Category A contains inventory that is most important to the business. The category usually
contains items that are carried in relatively low quantities but have a high individual value.
This inventory needs a high degree of attention. The business has a lot of money tied up in
the inventory but it carries low quantities and so could easily run out. The usage of this
category of inventory needs careful forecasting and the level should be closely monitored.
Category B inventory needs less sophisticated management than category A. This is an
intermediate category between A and C.
Category C contains inventory that has low value but is carried in high volumes. This
requires much less close management than category A. The business can apply the simplest
control possible to this area of inventory.

Expert view 7.3: ERP and ABC analysis


Many businesses operate enterprise resource planning (ERP) systems. These are integrated computer
software modules which support the different process areas of the business. This can include financial
accounting, management accounting, customer relationship management, project management, supply chain
management, manufacturing processes and human resource management. ABC inventory analysis is usually
built into these systems.

Just-in-time (JIT) inventory management


JIT is an approach to inventory management that has gained popularity in recent years as part of
a broader lean manufacturing approach. It is usually integrated into a materials requirement
planning (MRP) system.

Expert view 7.4: Lean manufacturing


Lean manufacturing is a management philosophy which was developed within the Japanese manufacturing
company Toyota. The focus of lean manufacturing is to ensure that any expenditure results in the creation of
value for the end customer. Any activity or expenditure which does not add value is seen as wasteful and
therefore a target for elimination. The approach involves removing unnecessary processes and material
movements, streamlining systems to remove idle time or unnecessary delays, and reducing inventory levels.

As with the other inventory management techniques we have already examined, the aim of JIT is
to minimize inventory costs. However, the JIT approach focuses primarily on minimizing
inventory holding costs through efficient monitoring of inventory usage and placing orders so
that they arrive only when needed. The goal is to minimize the amount of cash tied up in
inventory, thereby improving ROCE.
Successful operation of a JIT inventory system requires good forecasting of production
requirements, short and predictable lead times, and good coordination with suppliers. Good
communication both up and down the supply chain is critical. Implementing JIT inventory
management often involves integrating information systems with suppliers, and good
communication is critical. Although many businesses have successfully implemented JIT and
have increased efficiency and cost savings as a result, it is important to recognize that the system
also has drawbacks. By removing buffer stock the business has no emergency fallback in case of
disruption in the supply chain.

Advantages of JIT:

A JIT system eliminates the cost of holding inventory.


Inventory obsolescence is avoided.
Cash is freed up for investment in other activities.
Warehouse space can be freed up for other value-adding activities.
Production can be more flexible and responsive to customer requirements.

Disadvantages of JIT:

There is no ‘buffer stock’ so late delivery of inventory will cause disruption in production.
The business becomes vulnerable to disruptions in the supply chain caused by labour
strikes, transportation problems or information system problems.
It is more difficult to respond to unexpected fluctuations in demand.

Extract 7.6: JIT at Dell Computers


Dell, a leading supplier of personal computers, revolutionized the industry by using JIT systems. Dell takes
orders directly from a customer and builds computers to that customer’s order, an approach to production
which is known as a pull system. The company keeps only five days of inventory on hand, which compares
with competitors who have up to 90 days of inventory.
This approach has given Dell a competitive advantage over its rivals which enabled it to become the market
leader in personal computers. The reduced time to market enables Dell to keep up to date with technological
advancements so that customers are always offered the latest-specification computers. Dell has achieved this
by integrating the entire value chain from delivery and sales back to design and development.

Exercises: now attempt Exercise 7.4 on page 286

The financing of asset purchases


One of the greatest cash demands for businesses is the purchase of new assets such as machinery,
equipment or vehicles. Research has shown that the inability to finance the purchase of new
assets is one of the greatest barriers to growth for businesses.
If a business does not have the cash available to purchase new assets, it will need to raise
more cash by issuing new shares or by taking on a new loan. However, if the business is not able
to raise cash by these means, there are alternative methods of financing the purchase of new
assets which reduce the demands on cash flow.
One alternative method of financing assets is to use a hire purchase (HP) agreement. This
is a legal contract through which the purchases are made through a series of payments over time.
In reality the purchaser is hiring the asset and making a series of (typically monthly) rental
payments. Once these rental payments equal the cash purchase price plus an agreed interest
charge, ownership of the asset passes to the purchaser. If the purchaser defaults on the rental
instalments, the owner may repossess the asset. The main advantage of an HP agreement is that it
removes the need to have the purchase price in cash up front. The purchaser is able to spread the
cost of the asset over time through a series of payments.
Another method of financing the purchase of new assets is to lease those assets rather than
purchase them. There are two types of lease which a business could enter into: a finance lease or
an operating lease.
A lease is a contractual arrangement through which a party owning an asset (the lessor) gives
use of that asset to a second party (the lessee) in return for a series of rental payments.
Ownership of the asset remains with the lessor.
A finance lease is usually defined as one in which the risks and rewards of ownership of
the leased asset are transferred to the lessee but not the actual ownership. The lease term is all or
most of the life of the asset. The total rental payments throughout the lease period add up to the
cash price of the asset plus a finance charge. Responsibility for repair and maintenance of the
asset may remain with the lessor, but often this becomes the responsibility of the lessee. In many
cases, ownership of the asset is transferred to the lessee at the end of the lease term.
An operating lease is one in which the lease term is short in comparison to the life of the
asset. Operating leases are commonly used to acquire equipment or vehicles for short-term use.
For example, if you went on holiday and leased a car for a week, that would be an operating
lease. Under an operating lease the lessor remains responsible for maintenance and repair of the
asset, which can be an advantage for the lessee. However, this will be reflected in the lease cost.
If an asset is required infrequently for short periods, using an operating lease can be a smart way
of easing cash flows, as it removes the need to tie up valuable cash in an infrequently used asset.
On the other hand, it can be an expensive option if the asset is leased too frequently.
The decision of whether to enter into a finance lease or an operating lease will obviously
depend upon an assessment of how long an asset is required and how frequently it will be used.
A finance lease is essentially an alternative means of financing the acquisition of an asset. On the
other hand, an operating lease is a means of acquiring the use of an asset for short periods of time
without having to tie up cash in the purchase of that asset.
There are also tax and accounting disclosure implications for businesses deciding between
hire purchase, finance leasing and operating leasing. These may have an impact on the finance
method chosen.

Interpreting and analysing a cash-flow forecast


Now that you have learnt how to produce a simple cash-flow forecast, we will look at a more
complex example and learn how to analyse and interpret it. Within larger organizations it will
usually be the case that cash-flow forecasts are produced by the accounting department. As a
manager you will be using that forecast to make both operational and strategic decisions.

Expert view 7.5: Cash-flow forecasting


Large organizations usually have a specialist division within the accounting department that is dedicated to
cash planning and cash-flow forecasting. This is usually known as the treasury department.

Worked Example 7.9 sets out a cash-flow forecast for the start-up of a new business. We will
take a detailed look at this forecast and identify any potential problems which may be looming
for the business. We will also identify any changes to business plans that could be recommended
in order to improve the business’s cash flow and cash position.

WORKED EXAMPLE 7.9 Cash-flow analysis


Jill has recently been made redundant from her job as a designer. She has decided to use her redundancy money to
realize her dream of starting a business which manufactures and sells high-quality greenhouses. In order to start the
business Jill has applied for a business start-up loan from the bank. The bank requires a business plan, which
includes a cash-flow forecast. Jill has prepared the cash budget for the first six months of the business (Table 7.4).

Table 7.4

June July Aug Sept Oct Nov


Receipts $ $ $ $ $ $
Sales 0 0 40,000 60,000 60,000 80,000
Payments
Materials 0 30,000 20,000 20,000 35,000 50,000
Wages 3,500 3,500 3,500 4,200 4,200 4,200
Drawings 1,750 1,750 1,750 1,750 1,750 1,750
Workshop 900 900 900 900 900 900
Advertising 8,000 1,000 1,000
General 400 400 400 400 400 400
Van 12,000
Land Rover 450 450 450 450 450 450
Equipment 12,000 25,000
Loan 1,200 1,200 1,200 1,200 1,200
39,000 39,200 54,200 28,900 43,900 58,900
Increase/(Decrease) (39,000) (39,200) (14,200) 31,100 16,100 21,100
Opening Balance 55,000 16,000 (23,200) (37,400) (6,300) 9,800
Closing Balance 16,000 (23,200) (37,400) (6,300) 9,800 30,900

In relation to the cash budget, Jill has provided you with the following information:
1. The business will commence on 1 June with $55,000 in the bank. $30,000 of this will come from Jill’s
redundancy money. The remaining $25,000 will come from a business start-up loan.
2. Jill will rent a workshop unit on an industrial estate. This will cost $900 per month, payable at the start of each
month.
3. A van will be bought in June at a cost of $12,000.
4. General workshop costs (light, heat, power etc) and the van’s running costs are expected to be $400 per
month.
5. Jill will also lease a Land Rover Discovery car starting in June. The monthly lease payments will be $450,
payable on the 10th of each month.
6. The greenhouses will be sold to local garden centres. Sales are expected to be as follows:

The sales price of each greenhouse will be $4,000. In order to stimulate interest from customers, Jill has
offered two months’ credit.
7. Purchases of materials, which will be on one month’s credit, are expected to be as follows:

8. Jill will initially employ a workshop manager at a cost of $1,500 per month and two further staff at a cost of
$1,000 per month each, payable on the last day of the month. When sales reach 20 greenhouses per month,
she will employ an additional person at a cost of $700 per month. Jill will herself draw $1,750 cash each month
from the business.
9. Jill wishes to undertake a substantial advertising campaign to launch the business. This will cost $8,000 in the
first month and $1,000 per month for a further two months, payable in the month incurred.
10. The machinery and equipment needed will be bought immediately at the start of June. Most can be bought
second-hand for $12,000 cash. Other machinery, costing $25,000, will be bought new on two months’ interest-
free credit.
11. Repayments on the business loan, including the interest element, will be $1,200 per month. These will be
payable on the 15th of the month, commencing in July.

Required:
Analyse Jill’s cash budget and make any recommendations which you feel would improve her cash management.

Answer
The cash budget reveals that despite starting the business with an investment of $55,000, Jill will very quickly run out
of cash. In both of the first two months of business there is a predicted fall in cash of nearly $40,000. This means that
by the end of the second month of trading Jill will have to find another $23,200. By the end of the third month this
amount will have increased to $37,400. These forecast figures would suggest that unless Jill secures a substantial
additional source of finance (nearly $40,000), her business will fail within two months.
Alternatively, Jill can make changes to her business plans to reduce the cash requirements of the business. In this
way Jill may be able to launch her business successfully without the need for an extra $40,000.
We would suggest that Jill makes the following changes to her business plans:

Inventory management
The forecast shows that Jill intends to buy a substantial amount of raw material inventory in the first month. This
means that cash will be tied up in inventory that may not be realized as cash for several months. We recommend that
Jill buys materials as they are needed and not create a large inventory. She should seek out suppliers who are able
to supply quickly so that she can manage the business successfully with relatively low levels of inventory.

Management of accounts receivable


Jill is proposing to offer an extremely generous credit period to new customers. The impact of this is that there will be
no cash receipts for the business for the first two months. The business cannot afford such a generous credit policy.
Jill should therefore aim to gain orders from customers without the need for such a long credit period. If she can gain
sufficient customers in the first few months offering only a one-month credit period, this will increase the cash inflow
into the business by $40,000, substantially easing the cash-flow problem.
Jill should also consider using either invoice discounting or debt factoring to release some of the cash that will be
tied up in debtors.

Expenditure plans
Excessive expenditure can be a drain upon the cash flow of a business. Jill therefore needs to look at areas of her
business plan where she can cut expenditure to keep cash within the business.
One potential area is Jill’s own planned drawings from the business. Jill plans to draw $1,750 from the business
each month, right from the very start of the business. Jill should look at her own personal spending and reduce this
amount to the absolute minimum that she can manage to live on for the first few months of the business. Then, once
the business is well established and cash flows are eased, she will be able to draw more money in the future.
Another potential area for savings is the advertising plans. Jill plans to undertake a substantial advertising
campaign to launch the business. This will cost $8,000 in the first month and $1,000 per month for a further two
months. As Jill intends to deal with local garden centres she may be able to secure sufficient sales through personal
contact and negotiation. This will reduce or totally remove the need for the advertising campaign. Alternatively, Jill
can still advertise using cheaper methods.

Purchase and financing of assets


Substantial cash expenditure on new vehicles (the van and the Land Rover) and new equipment is planned. Jill
should look both at her needs and at the planned method of financing the purchases.
Jill plans to buy a van in the first month of the business at a cost of $12,000. This cost could be spread if Jill
purchased the van on a finance deal or lease. This would substantially ease the cash flow in the first few months of
the business.
Jill should also consider whether the Land Rover is necessary and if so, whether she needs to buy it so soon after
the launch of the business. Jill may aspire to driving a Land Rover but a cheaper vehicle may be sufficient for her
needs until the business is established and has more cash available.
Jill plans to buy a substantial amount of equipment ($37,000) within the first few months of the business. The cost
of this would be spread if the equipment could be bought on a finance deal. This may prove more expensive if the
equipment cannot be bought second-hand using a finance deal, but Jill needs to balance out the cash-flow benefits of
a finance deal with the reduced cost of second-hand equipment.
Repayments on the business loan are $1,200 per month. Jill could try to negotiate a longer period for the loan so
that the repayments could be spread over more months and thereby reduced in the critical first months. Alternatively,
she could ask for a deferment in the repayments so that she does not need to make any payments in the first two or
three months.

Management of accounts payable


Purchases will be made on one month’s credit. Jill should try to negotiate a longer credit period with her suppliers or
seek out alternative suppliers who are willing to offer more credit. In addition, she could seek early payment discounts
with suppliers.

Conclusion
In this chapter we have looked at the importance of cash management. We have explored the
reasons for holding cash and the different motives which determine the level of cash a business
will hold. We have examined models which aim to assist managers in optimizing cash levels
within the business. We have also explored how good management of different assets within the
business, such as inventory, accounts receivable, accounts payable and non-current assets, can
greatly improve the cash-flow requirements of the business. Lastly, we have explored the
importance of cash-flow planning and how managers can use cash-flow budgets to inform
planning decisions.

COMPREHENSION QUESTIONS

1. Explain the business problems that may be associated with holding insufficient cash
balances. What are the problems of holding too much cash?
2. Distinguish between the transaction motive and the speculative motive for holding
cash. How do these two motives differ in terms of their impact upon the level of cash
held?
3. Explain how a business can benefit from a cash forecast, even if it already has an
income and expenditure forecast and a balance sheet forecast.
4. Discuss the assumptions which underpin the Baumol–Tobin model of cash
management and explain how they may limit the usefulness of the model for some
businesses.
5. Explain the difference between a finance lease and an operating lease as a means of
financing the purchase of an asset.
6. Discuss the ways in which factoring and invoice discounting can assist in the
management of accounts receivable.
7. Explain what you understand by the terms ‘buffer stock’ and ‘lead time’ and briefly
consider any stock policy that would minimize or eliminate such costs.
Answers on pages 286–288

Exercises
Answers on pages 288–291

Exercise 7.1: Managing the cash account


The management of Darum Co have set a minimum cash balance of $7,500. The average cost to the company
of making deposits or selling investments is $24 per transaction. An analysis of cash flows over the last 12
months reveals a standard deviation of $2,000 per day. The average interest rate on investments is 4.6 per
cent.
Calculate the spread, the upper limit and the return point for the cash account of Darum Co using the Miller–
Orr model. Explain the relevance of these values for the cash management of the company.

Exercise 7.2: Cash-flow forecasting


Jade opens a sandwich shop. She sells sandwiches directly over the counter for cash, but also provides buffet
lunches for local businesses. These customers are allowed one month’s credit.
Budgeted sales for the first six months of the business are as follows:

Jan Feb Mar Apr May Jun


$ $ $ $ $ $
Counter sales 4,000 3,500 3,500 4,000 4,000 5,000
Buffet sales 4,000 5,000 3,000 4,000 3,500 4,000

Purchases of bread and fillings are all made locally for cash. Expected purchases are:

Jan Feb Mar Apr May Jun


$ $ $ $ $ $
Purchases 3,000 3,000 3,000 4,000 3,500 4,000

1. Jade employs two assistants in the shop. Each is paid a monthly salary of $1,000. This is payable on the
25th of the month.
2. Shop rent is $700 per month. This is paid on the first day of the month.
3. Shop heat and light costs are $300 per quarter, payable in arrears in March and June.
4. The bank balance on 1 January is $1,000.

Required:

(a) Prepare a cash-flow forecast for the first six months of Jade’s business.
(b) Comment on changes which Jade might make in order to improve her cash flow.

Exercise 7.3: Managing accounts receivable


Andro Co makes all its sales on credit and allows its customers 30 days’ credit. However, analysis of the
financial statements shows that the average accounts receivable period in the last financial year was 70 days.
This has increased from 50 days in the previous financial year. In addition, bad debts as a percentage of sales
increased from 3 to 6 per cent. The CEO has expressed great concern at these figures and has asked for your
assistance in improving the management of accounts receivable.

Required:
Write a report to the CEO of Andro Co which details ways in which the company could improve the
management of its accounts receivable.

Exercise 7.4: Inventory management


Rio Co currently employs an inventory management policy of reordering inventory when levels fall to 3,500
units. This reorder level allows for the lead time of two weeks and maintains a buffer stock.
The company orders 10,000 units at a time. Forecast production demand during the next year is 70,000
units. The cost of placing and processing an order is $300, while the annual cost of holding a unit in the
warehouse is $1.50. These costs are expected to be constant during the next year.

Required:

(a) Calculate the cost of the current ordering policy.


(b) Determine whether a saving could be made by using the EOQ model.

Answers to comprehension questions


1. If a business holds insufficient cash:
• Payments to suppliers may be delayed, causing credit problems and disruption in supply.
• Emergency loans or overdrafts, which cost the business more than is necessary, may be required.
• Opportunities for good investment may be missed.
• The business may ultimately become insolvent
If a business holds too much cash:
• ROCE will be reduced.
• The cost of capital may increase.
• The business may become a target for acquisition.
2. The transaction motive refers to holding cash in order to be able to pay day-to-day bills. The speculative
motive refers to holding cash in order to be able to take advantage of unexpected investment
opportunities. The amount of cash held due to the transaction motive will depend upon the cash
conversion cycle of the business. The longer this cycle (for example, due to offering extended credit to
customers), the more cash the business will need in order to continue with day-to-day transactions. The
amount of cash held due to the speculative motive will depend upon the attitude of management and the
investment opportunities available to the business. If interest rates are low, businesses will typically
maintain higher speculative balances of cash. On the other hand, if interest rates are high, the
opportunity cost of holding cash balances may outweigh the speculative benefits.
3. An income and expenditure forecast will enable a business to forecast its future levels of income and
profitability. A balance sheet forecast enables a business to forecast levels of assets and liabilities.
Although both of these forecasts are important to the business, neither provides a prediction of the cash
requirements of the business. A business needs to produce a cash budget in order to ensure that the
business holds sufficient to cash to meet the targets set within the income and expenditure budget. A
cash budget shows the amount and timing of cash receipts and cash payments. It helps a business
predict future cash needs and plan to meet any potential shortfalls.
4. The Baumol–Tobin model of cash management makes the following assumptions:
• Cash outflows are predictable and even over time.
• Cash inflows are predictable and regular.
• Day-to-day cash needs are funded from a bank current account.
• A ‘buffer’ of short-term investments is held which can be liquidated to provide cash when needed.
Although many businesses do have consistent cash flows, many have cash flows that are irregular or
unpredictable. If this is the case, a more useful model for cash management is one which allows for the
regularity and fluctuations in cash flows. One such model is the Miller–Orr model.
5. In a finance lease the risks and rewards of ownership of an asset is transferred to the lessee but not the
actual ownership. The lease term is all or most of the life of the asset. The total rental payments
throughout the lease period add up to the cash price of the asset plus a finance charge. Responsibility for
repair and maintenance of the asset may remain with the lessor, but often this becomes the responsibility
of the lessee. In many cases ownership of the asset is transferred to the lessee at the end of the lease
term.
In an operating lease the lease term is short in comparison to the life of the asset. The lessor remains
responsible for maintenance and repair of the asset.
6. Factoring and invoice discounting are two ways in which a business can release the cash tied up in
accounts receivable. Debt factoring involves a company handing over administration of its sales ledger to
a financial institution with expertise in this area. This may be a bank or a specialist factor company. The
factor will offer finance to the company, usually based on up to 80 per cent of the face value of invoices
raised. The loan is repaid as customers settle invoices, with the balance being passed to the company
after deduction of a fee equivalent to an interest charge on the cash advanced. Invoice discounting is an
alternative way of raising finance against accounts receivable. Rather than employing the credit
management and administration services of a factor, a number of invoices are offered as collateral
against borrowing. This approach can be considerably cheaper than factoring.
7. Buffer stock is a minimum level of inventory that a business holds in order to protect itself against
stockouts. Lead time refers to the time lag between placing an order for new inventory and actually
receiving it. A stock policy which seeks to minimize or eliminate the costs of buffer stock and lead time is
the JIT approach. This approach involves careful coordination with suppliers to ensure that new inventory
arrives just as it is needed.
Answers to exercises

Exercise 7.1: Managing the cash account


Spread: using the Miller–Orr model the spread can be calculated as follows:

Upper limit:
Lower limit (set by management) = $7,500
Upper limit = 7,500 + 24,036 =$31,536
Return point:
Return point = 7,500 + (24,036/ 3) = $15,512

Relevance for cash management


The Miller–Orr model takes account of uncertainty in relation to receipts and payment by setting limits within
which cash balances can be allowed to fluctuate. If cash balances fall to the lower limit, an amount of cash
equal to the difference between the return point and the lower limit is raised by selling short-term investments.
If cash balances rise to the upper limit, an amount of cash equal to the difference between the upper limit and
the return point is transferred into short-term investments. The model therefore helps Darum Co to decrease
the risk of running out of cash, while avoiding the opportunity cost of having unnecessarily high cash balances.

Exercise 7.2: Cash-flow forecasting


(a) Jade’s sandwich shop – cash flow forecast. See Table 7.5.

Table 7.5

Jan Feb Mar Apr May Jun


Cash sales 4,000 3,500 3,500 4,000 4,000 5,000
Credit sales 4,000 5,000 3,000 4,000 3,500
Total 4,000 7,500 8,500 7,000 8,000 8,500
Expenses
Purchases 3,000 3,000 3,000 4,000 3,500 4,000
Salaries 2,000 2,000 2,000 2,000 2,000 2,000
Shop rent 700 700 700 700 700 700
Heat & light 300 300
5,700 5,700 6,000 6,700 6,200 7,000
Net movement (1,700) 1,800 2,500 300 1,800 1,500
Opening balance 1,000 (700) 1,100 3,600 3,900 5,700
Closing balance (700) 1,100 3,600 3,900 5,700 7,200

(b) Improvements to cash flow: The current cash-flow forecast suggests that Jade will require additional
funding of $700 by the end of the first month of trading. She may be able to avoid this by taking one or
more of the following actions:

require a deposit or partial payment in advance from the buffet customers;


seek credit for purchases from some of her suppliers;
delay employment of a second assistant until February or March when she has more funds available.

Exercise 7.3: Managing accounts receivable


Report to CEO of Andro Co

Management of accounts receivable


The information made available to me concerning accounts receivable indicates that there are two areas of
concern:

the increasing level of bad debts as a percentage of credit sales; and


the excessive credit period being taken by credit customers.

Reducing bad debts


Bad debts have increased from 3 to 6 per cent of credit sales in the last year. This figure could be reduced by
introducing a credit policy which ensures that the creditworthiness of new customers is assessed before they
are extended credit. Andro Co needs to introduce a formal policy which sets out how this should be done. For
each new customer, information should be gathered from a variety of sources such as trade references, bank
references and credit reference agencies. The solvency and credit history of all new customers should be
evaluated. The terms of the credit offered will be based upon this assessment of default risk.

Reduction of average accounts receivable period


Customers have taken an average of 70 days’ credit over the last year rather than the 30 days offered by
Andro Co. As this is more than twice the formal credit period offered, there is clearly a problem in collecting
outstanding accounts receivable.
Andro Co should check the credit offered by its competitors. It may well be the case that competitors are
offering a longer credit period (for example, 60 days) and that Andro’s credit policy is therefore out of line with
the industry.
Andro Co should have a clear policy for managing outstanding accounts receivable balances. This should
include the following actions:

Statements should be regularly sent to customers detailing any amounts outstanding.


An aged accounts receivable analysis should be produced at the end of each month to identify customers
with overdue balances.
Any customers with outstanding accounts should be contacted to encourage payment.
Customers with overdue accounts should be denied further credit until those accounts are settled.
Customers with overdue accounts should be charged interest on those balances.

Andro Co could also consider introducing an early settlement discount as a positive means of encouraging
customers to settle their accounts sooner.

Exercise 7.4: Inventory management


Cost of current ordering policy:

Ordering cost = $300 × (70,000/10,000) = $2,100 per year


Weekly demand = 70,000/52 = 1,346 units per week
Consumption during 2 weeks’ lead time = 1,346 × 2 = 2,692 units
Buffer stock = reorder level less usage during lead time = 3,500 – 2,692 = 808 units
Average stock held during the year = 808 + (10,000/2) = 5,808 units
Holding cost = 5,808 × $1•50 = $8,712 per year
Total cost = ordering cost plus holding cost = $2,100 + $8,712 = $10,812 per year

Cost using the EOQ:

EOQ = ((2 × 300 × 70,000)/1•5)1/2 = 6,481 units


Number of orders per year = 70,000/6,481 = 11 per year
Ordering cost = $300 × 11 = $3,300 per year
Holding cost (including buffer stock) = $1•50 × (808 + (6,481/2)) = $6,073 per year
Total cost of EOQ-based ordering policy = $3,300 + $6,073 = $9,373 per year
Saving for Rio Co by using EOQ-based ordering policy = $10,812 – $9,373 = $1,439 per year
08
Pricing decisions

OBJECTIVE
To provide an understanding of the factors, both internal and external to the business, that should guide the pricing
decision.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Recommend appropriate costing methods for pricing.


Assess the potential impact of different competitive environments on pricing.
Identify appropriate pricing strategies to fit different markets and products/services.

KEY TOPICS COVERED


Costing and pricing – different approaches to costing.
Consumer behaviour and pricing.
Competitor behaviour and pricing.
Pricing new products or services.
Product life cycle and pricing.
Special pricing strategies.

MANAGEMENT ISSUES
Pricing is not an exact science. Much of the detailed mathematics of the economics of pricing is impossible to apply in
practice. Managers need an understanding of the broad principles which should guide them in the pricing decision.

Introduction
In this chapter we will examine the factors which a business needs to take into consideration
when establishing the most appropriate price for its products or services. In practice, pricing
strategies can be both short term and long term and businesses will employ a combination of
pricing strategies to achieve their corporate goals. For example, a business may have a long-term
goal of being a high-quality provider and will have a long-term strategy of setting prices high to
reflect this. However, the business may reduce prices in the short term in order to improve cash
flows, increase market share or deter competitors.
Some of the factors involved in pricing may appear obvious. For example, it is important to
take into consideration the costs of delivering a product or service in order to ensure that profit is
made or at least costs are covered. However, how those costs are established is not necessarily
always obvious or simple. Also, a business needs to be mindful of its customers and competitors
when establishing pricing strategy. How will customers and competitors react to your prices? In
some circumstances this may be an extremely important question, because if you get it wrong
you will lose all your customers to your competitors. The pricing decision can therefore be said
to be made up of 3 C’s: costs, customers and competitors (Figure 8.1).

Figure 8.1 The pricing decision

Another way of looking at this is to say that there are three different perspectives which need to
be considered when making a pricing decision:

The accountant’s perspective: this perspective is concerned with the relationship


between revenues and costs. The sales price must be set at such a level as to ensure that
revenues exceed costs in order to generate a profit.
The economist’s perspective: this perspective is concerned with the competitive
environment in which the business operates. Different market types have an impact on the
ability of a business to control its own pricing.
The marketer’s perspective: this perspective is concerned with the way in which
customers react to prices and how pricing should be integrated into the overall marketing
mix.
The accountant’s perspective can be said to be inward-looking in that it is concerned with the internal workings of the business
and how these generate costs. The economist’s perspective and the marketer’s perspective are both outward-looking in that
they are concerned with the business’s external environment and how that impacts upon price.

In this chapter we will look at each of these perspectives in turn to establish a good
understanding of the factors involved, and then we will see how all three perspectives can be
combined into coherent pricing strategies.

The accountant’s perspective – costing and pricing


Understanding the cost of delivering a product or service is important to all organizations. Profit-
seeking organizations have at least to break even and not-for-profit organizations need to cover
their costs.
The way in which a business will establish the cost of delivering its goods or services will
depend upon the way that the business operates and the information that is available. In this
section we will look at several well-established costing methods and the circumstances to which
they are most suited:

absorption costing and full-cost-plus pricing;


marginal-cost-plus pricing;
activity-based costing and pricing;
life-cycle costing and pricing.

Absorption costing and full-cost-plus pricing


Absorption costing, which is aimed at establishing the full cost of production, originally arose in
manufacturing businesses operating in a relatively stable environment. When using the
absorption costing approach, sales price is determined by calculating the full cost of a product or
service, including relevant overhead costs, and then adding a percentage mark-up to achieve the
desired level of profit.

WORKED EXAMPLE 8.1 Pricing using absorption costing


Grogstore Co manufactures stainless-steel beer barrels. The manufacturing depot has two production departments:
cutting and welding. The direct costs of producing each barrel are as follows:

Materials: 2 m2 of stainless steel @ $3 per m2


Labour – cutting: 15 minutes per barrel @ $14 per hour
Labour – welding: 20 minutes per barrel @ $18 per hour

Budgeted overhead costs for next year are:

$
Property costs 92,000
Managers’ salaries:
Cutting department 25,000
Welding department 28,000
General administration costs 143,000
Machine power 28,000

The following information relates to each of the production departments:

What price should Grogstore charge for the barrels if it wishes to earn a profit mark-up on total costs of 30 per cent?

Solution
The first step in establishing a full-cost price is to identify all the direct costs of manufacturing a barrel. These will
include the costs of materials and labour which go directly into the barrel:

Once total direct costs have been established, the relevant parts of overhead costs are incorporated into the product
cost. This is done through a method known as the ‘3 A’s’ of absorption costing: allocate, apportion and absorb.
This method involves attributing overhead costs firstly to relevant production cost centres, and then from the cost
centre to the individual product.

Expert view 8.1: Cost centre


A cost centre is a department or segment of the business to which costs are allocated or apportioned.

In the case of Grogstore Co we can identify two production cost centres: the cutting department and the welding
department. This is, of course, a very simple example and in reality most organizations will have many cost centres.
All overhead costs which can be directly attributed to a given cost centre should be allocated to it. For example, as
there is a manager responsible for the running of each department, his or her salary can be allocated to that
department.

Overhead costs which do not relate directly to one cost centre but rather relate to the running of the business as a
whole are then apportioned to each cost centre in the most appropriate and fair way. For example, the property costs
can be apportioned based upon space occupied by each cost centre:

In a similar manner, administration costs can be apportioned between the two production cost centres based upon
the number of employees in each department, and the machine power cost can be apportioned based upon the
number of hours worked in each department (labour hours). This will result in the total cost allocation and
apportionment between the two production cost centres shown in Table 8.1.

Table 8.1
Cutting Welding Total
$ $ $
Allocated costs:
Managers’ salaries 25,000 28,000 53,000
Apportioned costs:
Property costs (floor space) 53,667 38,333 92,000
Admin costs (no. of employees) 53,625 89,375 143,000
Machine power (hours worked) 9,692 18,308 28,000
Total 141,984 174,016 316,000

There is no fixed basis for apportionment of costs. The method used will depend upon perceptions of what is most
appropriate and what information is available. The most important factor is fairness – the allocation and
apportionment of costs must fairly reflect the extent to which the cost centre contributes to the generation of those
costs.
Once all overhead costs have been allocated or apportioned to production departments and a total cost calculated
for each department, that cost is ‘absorbed’ into individual products by calculating an overhead absorption rate
(OAR). This OAR is usually based upon the level of activity (or amount of time) contributed from each department in
the making of a single product.
Therefore, for the cutting department of Grogstore Co the OAR can be calculated based upon labour hours as
follows:

This means that for every hour of work done in the cutting department, $15.78 will be charged to the product. As each
barrel requires 15 minutes of cutting department time, it will be charged $3.95 ($15.78 × 15/60 minutes) of cutting
department costs.
The OAR for the welding department will be:

As each barrel requires 20 minutes of welding department time, it will be charged $3.41 ($10.24 × 20/60 minutes) of
welding department costs.
We can now establish the total cost of producing each barrel, including the relevant overhead costs (Table 8.2).

Table 8.2

Direct costs: $
Materials: (2m2 @ $3 per m2) 6.00
Labour – cutting: (15 minutes @ $14 per hour) 3.50
Labour – welding: (20 minutes @ $18 per hour) 6.00
Overhead costs:
Cutting department (15 minutes @ $15.78 per hour) 3.95
Welding department (20 minutes @ $10.24 per hour) 3.41
Total cost: $22.86

An appropriate sales price can now be calculated by adding the required profit mark-up of 30 per cent:

Sales price per barrel = $22.86 × 130% = $29.72


Expert view 8.2: Mark-up and margin
Once a full cost has been established for a product, the price is calculated by adding mark-up to that full cost.
This mark-up is usually a standard percentage which is established by management. For example, a company
may decide to add a mark-up of 25 per cent to all its products. If product A has a full cost of production of £80,
the sales price will be calculated as:

Sales price of Product A = £80 × 1.25 = £100

At this point it is worth discussing the difference between mark-up and margin, as these are two terms which
are frequently confused: mark-up is profit expressed as a percentage of costs; whereas margin is profit
expressed as a percentage of sales price.
A simple illustration will show the significance of understanding the difference. Imagine a product that has a
cost of £60 and a sales price of £80. This means that the profit will be £20.

Problems with full-cost-plus pricing


Although full-cost pricing is in theory a method of always ensuring that all costs are covered and
a known profit is earned, there are a number of problems with implementing this method in
practice.
Firstly, the technique requires the estimation of some important figures, such as annual
overhead costs and the volume of output. These figures will not be known for certain until an
accounting period is complete, yet they are needed at the start of an accounting period if a price
is to be established. Although the technique theoretically establishes the total cost of a product,
in practice this is just an estimation and not the actual cost. There is therefore always a risk that
estimates are inaccurate and the method does not establish an accurate cost for pricing.
Another problem is that the basis of apportioning and absorbing overheads may well be
arbitrary and may result in an unfair or appropriate apportionment of overhead costs between
different products and services. This may result in some products or services being overpriced,
which could make them uncompetitive. For example, one of the authors once worked in a
business division which operated from the old premises of the business after the other divisions
had moved into new premises. This building was far too big for the needs of the division and was
half empty. Despite this, the division was charged for the full running cost of the premises.
A further disadvantage is that by simply calculating the full cost as it currently stands and
adding a mark-up to this, there is no incentive within the system to reduce costs. When we look
at pricing within the competitive environment later in this chapter, it will be seen that this can be
a significant issue.
Marginal-cost-plus pricing
One way of overcoming some of the problems of full absorption cost pricing is to avoid the
process of estimating and apportioning overhead costs. Also, for some businesses, the basic
variable costs of production are easy to identify, but overheads are less easy to apportion and
allocate. In this case, sales price is determined by adding a profit mark-up to the marginal
(variable) cost. This method of costing is widely used in retailing and in professional services.
A retailer can easily identify the cost of products bought in for resale and can establish a
consistent gross profit margin by applying a mark-up to this cost. Different retailing businesses
have different levels of mark-up. In clothing retailing it is usual to have mark-up on purchase
cost of 400–500 per cent. However, in food retailing where volumes of sales are higher, profit
margins are much lower and mark-up on direct costs may be as little as 25 per cent.
The marginal-cost-plus pricing method is also frequently used by professional service firms
such as solicitors and accountants. For such businesses the direct cost of delivering a service will
be the salary of the solicitor or accountant working for the client. The price charged to the client
will be based upon a multiple of that salary. This is illustrated in Worked Example 8.2.

WORKED EXAMPLE 8.2 Marginal-cost-plus pricing


A solicitor is paid a salary of £30 per hour. For every hour which she spends working for a given client, that client will
be charged £30 × 300% = £90. This price will cover the direct cost of the solicitor’s salary and any office overhead
costs, and will leave a profit for the firm.

Exercises: now attempt Exercise 8.3 on page 323

Activity-based costing (ABC) pricing


Activity-based costing is a method of costing which was developed to address some of the
criticisms of absorption costing discussed above. In a bid to identify and apportion costs more
accurately, this approach attributes costs to different organizational activities, rather than
departments. The method therefore uses the concept of cost pools – the collection of costs
which relate to a given activity.
Under absorption costing, different overhead costs will be apportioned to a department which
may undertake a range of activities. For example, the production department may accumulate
costs for ordering, receiving and holding inventory, machining, assembly, packaging and
dispatching. Under ABC these would be identified as separate activities with separate cost pools.
The difference between the two approaches to costing is illustrated in Figure 8.2.
Figure 8.2 Traditional absorption costing vs ABC. (a) Traditional absorption costing
system; (b) activity-based costing system

Once costs have been apportioned to activities, the cost driver for each activity is
determined. The cost driver is the metric which best explains why resources are consumed by a
particular activity and therefore why the activity incurs costs. The cost driver provides an
explanation of the size of the cost pool. This system of activity-based costing is best illustrated
with an example which compares the approach with that of traditional absorption costing.

WORKED EXAMPLE 8.3 Activity-based costing and pricing


Pedal Co manufactures a range of different bicycles, including the A-series and the B-series. The A-series is a bicycle
aimed at serious sports riders. The company makes and sells 200 of these each year. The B-series is a leisure
bicycle. The company makes and sells 5,000 of these each year.
The bicycle frames are made in the company’s factory, but all other components, including wheels, are bought in
from other suppliers. The factory has a normal production capacity of 10,000 direct labour hours each year.
Production overhead costs relating to receiving raw materials and bought-in components are $240,000 per year.
These costs have been identified as relating to the following activities:

Receiving 225 consignments of frame tubing $90,000


Receiving 250 consignments of components $150,000

The A-series requires 60 frame-tubing consignments and 50 component assignments. The B-series requires 50
frame-tubing consignments and 75 component assignments.
The direct costs of producing each model of bicycle are as follows:
The company applies a standard mark-up on its products of 40 per cent.

Required:
Compare the total cost and therefore the price of the two models of bicycle using traditional absorption costing and
activity-based costing.

Answer: Traditional absorption costing


If Pedal Co uses traditional absorption costing, the $240,000 production overhead costs would be absorbed into each
model of bicycle using an OAR based upon production time:

Therefore the total cost and price of each model would be:

Answer: Activity-based costing


If activity-based costing is used, the overhead costs will be apportioned based upon cost drivers. In this case the cost
drivers are receiving the frame tubing and receiving the components. The absorption rate based upon these cost
drivers will be:
Receiving frame tubing = $90,000 ÷ 225 = $400 per consignment
Receiving components = $150,000 ÷ 250 = $600 per consignment
These costs will be apportioned to the two models of bicycle based upon the level of activity which each model
generates:

This ABC analysis suggests that more overhead costs should be apportioned to the A-series and less to the B-series.
This in turn has a knock-on effect on the prices which the company should be charging for each model of bicycle.
Advocates of ABC argue that this calculation provides a much more accurate analysis of the true cost of producing
each model of bicycle because the technique is more sensitive to how activities give rise to the overhead costs.

Life-cycle costing and pricing


With many modern high-technology products, life cycles have become much shorter and more
costs are incurred at a pre-production stage in research, development and design. This means that
direct production costs can be a relatively small part of the overall cost of the product. In such
cases it is more useful to identify the costs across the whole life cycle of the product and use
these as the basis of establishing a suitable price.
All products can be seen as having a life cycle, from development through to withdrawal from
the market. This life cycle, as illustrated in Figure 8.3, has different stages during which costs
and revenues will have differing patterns. During development, costs will be incurred but no
revenues are earned yet. At the introduction stage, sales may be slow to pick up, meaning that
initial sales revenue will be low. At the same time, costs may be relatively high because there is
insufficient sales volume to generate economies of scale. In the growth stage, sales begin to pick
up and grow rapidly and the product becomes more profitable. During the maturity stage, sales
will stabilize at a maximum level. This is usually the most profitable period for a product. As the
product enters its final stage of decline, sales and profitability will fall off until the point where
management are forced to withdraw the product from the market.

Figure 8.3 Life-cycle pricing

Life-cycle costing recognizes the fact that in order to make a profit, revenues must cover all
costs whether they are incurred pre-production, during production or post-production. This focus
on costs across the whole life cycle has many advantages. Firstly, focusing on all costs will help
management see opportunities for cost reductions. Secondly, many costs will be linked across
the life cycle and therefore opportunities for reducing costs later can be identified at an earlier
stage. For example, better design may reduce production costs and changes in manufacturing
processes may reduce end-of-life decommissioning costs.
A further dimension of life-cycle pricing is the recognition that different pricing strategies
may be adopted at different stages in the product’s life cycle. The best choice of pricing strategy
at any given stage will depend upon customers’ perceptions, competition and the nature of the
product being sold. All of these factors, together with the appropriate strategies, will be
discussed as we progress through this chapter.

Extract 8.1: Sony PS3 life-cycle costing


When Sony launched its PlayStation3 it was priced at $499, but it was costing Sony $806 per unit to make.
This represented a loss of $307 per machine. Sony launched the PS3 at this price because it expected to
reduce its losses over the 5–10-year life cycle of the machine as economies of scale reduced the cost of parts.
In fact, over the whole life cycle of the PS3 Sony expected to make a small profit. In addition, the bulk of Sony’s
revenue comes from selling or licensing software to run on the PS3. The pricing of the gaming platform then
needs to be competitive in relation to the main alternatives – Microsoft’s Xbox 360 and the Nintendo Wii.
Sony’s pricing decision was made after considering the full life cycle of both the gaming machine and the
software.

Conclusions: costing for pricing


A major disadvantage of all the costing methods examined above is that they are entirely inward
looking. That is to say, they ignore the impact which price has on demand. By simply calculating
the costs of production and adding a mark-up for profit, one is assuming that the customer will
be willing to pay the price that is arrived at through these calculations. In reality, customers are
not concerned with how much it cost to make a product. Rather, they are interested in the value
the product provides them and they will be reluctant to pay a price which does not offer good
value. Furthermore, the internal focus ignores the activities of competitors. If competitors are
producing similar products at lower costs, they will be able to undercut the business’s price. It is
therefore important to consider these external factors and these are the subject of the following
sections.

The economist’s perspective


From the economist’s perspective we look at the competitive environment of the business and
how this will impact on pricing strategy. There are two aspects of economic theory which we
want to consider in relation to pricing. The first is the concept of the price elasticity of demand.
The second concept is that of market type or market context: the competitive conditions in which
the business operates.

Price elasticity of demand


Generally speaking, as the price of a product or service increases, the demand for it will
decrease. However, the amount by which that demand decreases will vary across different
products or services, for a number of reasons (which will be explored later in the chapter). The
relationship of this change in demand to a change in price is known as the price elasticity of
demand (PED). Economists measure this with the following formula:

The greater the percentage change in demand in relation to a given price change, the more
‘elastic’ that demand is said to be. If a 10 per cent increase in price creates a 30 per cent fall in
demand, demand is said to be relatively elastic. On the other hand, if a 10 per cent increase in
price creates only a 5 per cent fall in demand, demand is said to be relatively inelastic.
It is important for a business to understand how elastic the demand is for its products or
services, as this price elasticity of demand will have an impact on the success of different price
strategies. For example, if a business knows that the demand for its products is relatively
inelastic, it would not pursue a strategy of cutting prices in order to try to stimulate demand. The
importance of this relationship is explained in the following section.

Price elasticity of demand and total revenue


In order to appreciate the importance of price elasticity of demand upon pricing strategies, it is
necessary to understand the relationship between demand, price and total revenue. In Figure 8.4
total revenue is represented by the area enclosed by the lines P1, A, D1. This area is a graphical
representation of the fact that price × demand = revenue:

Figure 8.4 Total revenue

Now look at Figure 8.5 where the price is reduced from P1 to P2. That reduction in price
stimulates an increase in demand from D1 to D2. In this example, demand is relatively inelastic
in relation to price so that the decrease in price creates a relatively small increase in demand.
Now look at the relative change in total revenue as indicated by the shaded boxes. It can be seen
that the decrease in revenue resulting from the price cuts (box P1–P2–A) is much greater than the
increase in revenue resulting from the increase in demand (box D1–D2–B). Therefore, overall
revenue falls as a result of the cut in price, even though demand has increased.

Figure 8.5 Price inelasticity


On the other hand, in Figure 8.6 a relatively small decrease in price creates a large increase in
demand. In this case the loss of revenue caused by the fall in price (box P1–P2–A) is more than
offset by the increase in revenue from the greater demand (box D1–D2–B). As a result, overall
revenue will increase due to the price cut.

Figure 8.6 Price elasticity

What determines elasticity?


The graphs above serve to illustrate the way in which the price elasticity of demand can be a
hugely important factor when it comes to pricing. Therefore, it is important for a business to
understand what determines the elasticity of demand for its products or services. Managers also
need to understand how they can take measures to change the price elasticity of demand.
Generally speaking, it is more beneficial for a business to have low price elasticity of demand.
This means that managers will seek methods for reducing elasticity. In this section we will look
at the factors which determine the price elasticity of demand and look at how it may be possible
to manipulate those.

The relative price of the goods


The more expensive that something is, relatively speaking, the more elastic demand will be. For
example, if a bar of chocolate costs $0.50 and the price increases by 20 per cent to $0.60, this
increase is unlikely to have a major impact upon demand. On the other hand, if a car costing
$25,000 increases in price by 20 per cent to $30,000, this increase will decrease demand as it will
put the car out of the affordable price range of some existing customers.
The price of other goods
Economists identify two different categories of related goods or services, the price of which can
have an impact upon the price of other goods or services.
The first category is complementary goods. These are products which go together, such as
video games and gaming consoles. If the price of a complementary good increases, this may
have a negative impact upon the demand of your product. For example, if the price of gaming
consoles increases so that fewer people can afford them, then fewer people will be buying games
(see Extract 8.1).
The second category is substitute goods. These are goods which a customer could buy instead
of the item we are trying to price because they can be substituted in its place. An example is two
mobile phones from different manufacturers with similar specifications. If the price of a
competitor’s phone is less than the price of your similar model, a customer will buy the
competitor’s phone as it represents better value.

Consumer income
Products or services can be categorized in terms of how their price elasticity of demand changes
with levels of consumer income. Some items, known as staple goods, will have a relatively
stable level of demand, regardless of consumer income. A simple example might be a basic
foodstuff such as potatoes or rice. Generally speaking, people tend to eat the same amount of
potatoes or rice, regardless of their level of income. An increase in income will not increase
demand for this type of goods.
A second category of goods, known as luxury goods, will be subject to an increase in demand
as consumers’ income increases. This will include items such as expensive holidays, luxury cars
and jewellery.
A third category of goods, known as inferior goods, will be subject to a fall in demand as
consumers’ income increases. These goods may not actually be inferior in terms of quality, but
they are in terms of consumers’ perception. Therefore, for example, a consumer on a low income
may buy a pair of cheap unbranded denim jeans. If that consumer has an increase in income, he
or she may stop buying the cheap jeans in favour of more expensive brand-name clothing.

Tastes and fashions


Tastes and fashions can have a significant impact on the elasticity of demand for products and
services. It is not unusual to see the latest ‘must have’ children’s toy, usually linked to a recently
released film, selling at a very high price and yet parents are falling over each other to buy it. On
the other hand, last season’s clothes can usually be found in the bargain bin when nobody wants
to buy them, even at a fraction of their original price.

Expectations
If customers expect the price of a product to increase in the near future, this can increase demand
in the short term. An example of this is seen in the UK each year before the Chancellor’s budget.
The usual expectation is that the Chancellor will increase taxation on vehicle fuel. This leads to
large queues at filling stations the evening before the budget, as drivers rush to fill their vehicles
in expectation of an increase in fuel prices.

Obsolescence
Aside from changes in taste and fashions, goods can also become obsolete. This is particularly
the case with high-technology products. Once a new version with a higher specification has been
released, demand for the older version usually falls off very quickly and customers are not
willing to buy that product even at a substantially reduced price. Another issue, related to
obsolescence, is how long a product lasts. If a product has a relatively short life before it wears
out or breaks, then over a long period customers will buy more of that item. However, this must
always be balanced with customer expectations of value. If a product wears out or breaks sooner
than the customer thinks it ought to, the customer may boycott that item in the future.

Pricing within market context


The second aspect of economic theory that is important to pricing is market type, or market
context. This refers to the competitive conditions in which the business operates. This can refer
to the number and type of competitors and the ability of a business to differentiate itself from
competitors.
Economists identity four types of market:

perfect competition;
monopoly;
monopolistic competition;
oligopoly.

Each of these different types of market has an implication for the business’s ability to determine
its own pricing strategy, and therefore whether it is a price setter or a price taker. It is therefore
extremely important for a business to understand what type of market it operates in and how this
will impact on it.

Perfect competition
Perfect competition describes the situation in which there are many buyers and sellers in a
market, none of whom is big or powerful enough to influence the market. New sellers can freely
enter the market at any time and all firms operating within the market are selling products or
services which are difficult to differentiate, so that they are essentially identical. This means that
customers are likely to choose between the products or services of different businesses purely
upon the basis of price. The consequence is that an individual business will find it very difficult
to set a price which differs from that of every other seller in the market. This situation is
illustrated in Worked Example 8.4.
WORKED EXAMPLE 8.4 Perfect competition and pricing
Imagine a large fruit and vegetable market in which there are 30 stalls all selling oranges which were purchased from
the same wholesaler and are therefore indistinguishable in quality. If 29 of the stalls sell the oranges at $0.50 each,
what will happen if one stall attempts to sell the oranges at $0.60 each? The likely outcome is that the more
expensive stall will be unable to sell any oranges, as customers, freely able to see the price and quality of the
oranges at all the other stalls, will buy oranges elsewhere. On the other hand, what would happen if one stall dropped
its price to $0.40 per orange? Again, because customers are able to compare prices and quality freely, they will all
rush to the one stall to buy the cheaper oranges. This will force all of the other stallholders to reduce their price to
$0.40. The ultimate outcome is that the original price-cutting stall will have no price advantage and all stallholders will
have lost profit by selling their oranges at a lower price.
The implication is therefore that the closer to perfect competition the market conditions are, the more difficult it will
be for an individual firm to set its own prices. Each firm selling within that market will be forced to go along with the
prevailing market price.
It is interesting that the internet and, in particular, price comparison sites have made information much more freely
available to customers and have created a situation very close to perfect competition for many retailers.

Monopoly
A monopoly describes a situation where a market is dominated by one major seller. In this
situation customers will have little or no option to buy from an alternative supplier. This gives
the monopoly business far more control over pricing. A business can have a monopoly through
ownership of patents or copyrights or through access to limited resources. Any aspect of the
market which creates a barrier to entry for other firms can result in a monopoly.
Provided that there is not a very high level of elasticity of demand, a monopoly will be able to
charge higher prices and customers will continue to buy from them, simply because they have no
other option. Because this creates a situation in which the supplier can potentially abuse
customers, many jurisdictions have created competition laws to prevent or restrict monopolies. In
the UK many industries have been broken up to remove monopolies in an attempt to create more
price competition among suppliers. Examples include electricity and gas suppliers,
telecommunications and rail transport.

Monopolistic competition
Monopolistic competition describes a market in which there are multiple buyers and sellers, but
sellers are able to differentiate their products or services from those of their competitors. This
means that if you are able to persuade customers that your product is somehow better than that of
your competitors, for example by having better features or being of a better quality, you will be
able to charge a higher price.
The market for many common products and services can be seen as falling into this category.
Businesses operating in such a market tend to spend a lot of money on marketing to establish
brand loyalty and to convince customers that their products should be purchased in preference to
those of competitors.

Oligopoly
Oligopoly describes the situation where a market is dominated by a small number of suppliers
selling similar products or services. A surprisingly large number of markets are oligopolies: car
manufacturers, computer manufacturers, petrol and oil suppliers, gas and electricity suppliers and
mobile telephone providers are all oligopolies. When there is an oligopoly, even though there are
multiple suppliers in competition, the suppliers are few enough in number to be able to cooperate
and work together to control prices. An example of this is OPEC, the group of oil producing
nations that work together to control oil prices.

Extract 8.2: The perils of trading with an oligopoly


The supermarket industry in the UK is a well-known oligopoly, being dominated by four major businesses:
Tesco, Sainsbury, Asda and Morrisons. This means that leading food producers must trade with the
supermarkets, which consequently dominate their order book and squeeze their profit margins. However,
smaller food producers are often unable to deal with the major supermarkets and therefore sell to independent
retailers. As this is a more open market, it enables them to charge more reasonable prices and earn higher
profit margins.

The marketer’s perspective


The marketer’s perspective on price is concerned with the impact that price will have upon
consumers’ perception and purchasing behaviour. Price is therefore one of the 4 P’s of the
marketing mix, the others being product, promotion and place.

Expert view 8.3: the marketing mix


The marketing mix is that set of factors which are regarded as being crucial in marketing a product or service.
Since the 1960s this has been commonly referred to as the 4P’s: price, product, promotion and place
(McCarthy, 1960). However, this model has been refined and developed over time so that there is now a 7P’s
model of the marketing mix (price, product, promotion, place, people, process and physical evidence) which is
considered to be more appropriate for marketing services rather than products. Also, more recently a 4C’s
model has been proposed as being a more consumer-focused view of the marketing mix. This has two
alternative versions, the first being consumer, cost, communication, convenience (Schulz et al, 1993) and the
second being commodity, cost, communication, channel (Shimizu 1973). Regardless of the variation in detail
and emphasis, in all these models of the marketing mix the price or cost to the consumer is highlighted as
being a key factor.

The impact of price on consumer behaviour


From a marketing perspective, price is important because it has an impact on the customer’s
perceived value of the product. It is therefore not so much the price itself but the customer’s
perception of that price which is important. If customers can be convinced that a higher price
represents good value for a product or service, they will be willing to pay a higher price.
In the section on the economist’s perspective we looked at a number of factors which can
affect the price elasticity of demand. The marketer is equally concerned with these factors, as
good marketing may be able to reduce the price elasticity of demand. That will enable the
business either to charge higher prices or to gain more customers at existing prices.

Expert view 8.4: Product positioning


In a competitive market it is important to make your products or services stand out from those of your
competitors, an activity which is known as product positioning. There are several academic models of product
positioning which look at different aspects of what is important to customers, but pricing is widely recognized as
being key. Generally speaking, businesses will want to charge the highest price they can, but their ability to
persuade customers to pay a higher price will depend upon customers’ perception of the quality of what they
receive for their money. The model in Figure 8.7 illustrates how price maps against quality in product
positioning.

Figure 8.7 Product positioning

The product positioning price/quality model shows four general strategies which a business might follow (A,
B, C or D). Of the four quadrants available, only quadrants C and B offer sensible strategies. Strategy A is to
charge a lower price for a high-quality product. This strategy will certainly attract customers but it will not be
one which maximizes profit for the firm. Strategy B is to charge a high price for a high-quality product. This is a
strategy followed by many businesses that sell luxury products. Customers are willing to pay a higher price
because they believe that what they are getting is a better product. Strategy C is to sell a lower-quality product
at a low price. This is another common strategy. The business will charge a low price and thereby gain a high
volume of sales, so that even if the profit per item is low the overall profitability will be high. Strategy D is to sell
a low-quality product at a high price. This strategy is unlikely to succeed in the long term as customers will
realize that they can get better-quality products at the same price elsewhere.
Many businesses combine strategies B and C by offering a range of products at different qualities and
different prices. For example, most major supermarkets have a budget low-cost range, a normal range and a
luxury high-cost range of foods.

Price perception
Customers are looking for good value from any product or service they buy and will therefore be
unwilling to pay a price which they believe does not offer value. The level of sensitivity to price
will generally vary with the absolute price being charged. That is to say, most customers will
unhesitatingly buy something which is relatively cheap, say $0.50, but will stop and think before
buying something which is more expensive, such as $1,000. (As explained in the economist’s
perspective above, this will also vary according to the level of the consumer’s disposable
income.)
If customers are looking for value from a product, it can be useful to make customers believe
that a product is cheaper than it actually is. It is therefore not unusual to see products priced at
$9.99 rather than $10, or $79.99 rather than $80. The objective of this pricing, which is
sometimes known as psychological pricing, is to overcome any perceived price barriers that
customers may have. A customer may consider $10 to be too expensive, so pricing the product at
$9.99 makes an important psychological difference.

Perception of quality
If customers perceive a product or service to be of a higher quality, they will be willing to pay a
higher price. It is the customer’s perception of quality rather than the actual quality itself which
is important. This is why many businesses invest in heavy advertising which emphasizes the
quality of their products.

Perception of difference from other products


Differentiating your products from those of competitors is another way of persuading customers
to pay a higher price. This involves persuading customers to buy based upon other elements of
the marketing mix, such as special or unique features of the goods, special promotion activities,
brand loyalty, and place (where the goods or services can be obtained).

Consumer ethics
Goods or services which satisfy customers’ ethical concerns can command higher prices. For
example, customers will be willing to pay higher prices for free-range eggs, fair-trade products,
carbon-neutral products, sustainable products or foodstuffs which have been farmed organically
in order to reduce damage to the environment.

Differentiating customers
Another important aspect of marketing is the recognition that not all customers will behave in the
same way. For a given product within a given market, some customers will have relatively high
price elasticity of demand whereas other customers will have relatively low price elasticity of
demand. An important skill in marketing is identifying and targeting different groups of
customers according to their willingness to pay. An example of this is supermarkets which have
a ‘value’ range of products for thrifty customers, a normal range of products for the majority of
customers, and a premium or luxury range of products for better-off customers.
Combining the three perspectives: establishing an
appropriate pricing strategy
Now that we have looked at the different issues which go into the pricing decision and the
factors which influence prices, we can look at how organizations should consider these when
establishing pricing strategies. Figure 8.8 sets out a range of different pricing strategies. Before
we look at these in detail, we need to consider some general principles. Firstly, the long-term
pricing strategy of an organization should be in line with the organization’s mission and broader
strategic objectives. Secondly, any pricing strategy adopted should always be informed by all
three C’s of pricing discussed above: costs, competitors and customers.
Broadly speaking, there are three considerations that will help the business frame an
appropriate pricing strategy. These relate to the following questions:

1. Are we pricing a new or existing product or service?


2. Are we going to a new or existing market?
3. Are we a volume-driven or a price-driven business?

New product or existing product?


If a business is launching a new product or a product with substantial new features (ie one that is
not currently offered by any competitor), there will be no competition for that product. This may
mean a relatively inelastic demand and a high level of interest in the product. On the other hand,
if the business is launching a product which is similar to those already available from
competitors, it needs to compare that new product with those already available. Are the features
comparable? Are there new or unique features? These factors will impact upon the business’s
ability to charge higher prices. For example, when Apple launched the iPad there were no
comparable products available. This, together with the relative price inelasticity of the market,
enabled Apple to charge high prices for their new product. However, as other competing
manufacturers launched similar products they had to price more competitively in recognition of
the fact that there were several substitute products available.

New market or existing market?


If a business is entering an existing market where other similar products are already on sale, they
will need to analyse what competitors are charging and set their prices accordingly. On the other
hand, if a business is entering a new market where there is no existing competition, pricing will
be based upon expectations of what customers are willing to pay.

Volume driven or price driven?


Businesses that have large fixed overhead costs tend to be volume driven in their pricing
strategy. For example, an airline will have a largely fixed cost of operating a flight. It will
therefore manipulate its prices in order to fill all the seats on an aeroplane whilst covering those
fixed costs. In a similar manner, supermarkets have high fixed costs. They price to sell a high
volume of products at a low margin. On the other hand, businesses that are price driven tend to
have low fixed overheads. Such businesses do not need high volumes of sales.

Pricing strategies
Having considered these overriding issues, let us look at the range of pricing strategies available.
These can be seen as a range of pricing positions in relation to the prices of competitors, as
illustrated in Figure 8.8.

Figure 8.8 Price-positioning strategies

1) Market skimming pricing


A market skimming pricing strategy involves launching a new product at a very high price and
then gradually reducing the price over time. This strategy is often used to launch a new
technology product which has little or no competition. It takes advantage of the fact that demand
for the new product will be relatively inelastic for some customers. This strategy is effectively a
type of price discrimination which segments the market in terms of how long customers are
willing to wait before obtaining the new product. For example, if a new specification of smart
phone is launched, some customers, referred to as early adopters, will want to buy that new
phone immediately, with little regard to the cost. The seller is therefore able to charge a very
high price at launch. On the other hand, some consumers will only buy that new phone once the
price is substantially lower, and they are willing to wait several months after launch before
buying it.
Price skimming is by necessity a short-term strategy applicable only to the launch stage of a
new product, as it relies on the uniqueness and novelty of the new product. As competitors
launch comparative products this will put downward pressure on prices. A business also has to
be careful in the timing of reducing the price. If prices are kept high for too long, a relatively
small volume of sales will be achieved and there is a risk that a competitor may develop an
alternative product which goes on to take a higher volume of the market.

2) Optional pricing
Optional pricing is another method of segmenting the market in order to maximize profit. This
strategy is used by many car manufacturers. Rather than just offering one version of a particular
model of car, the manufacturer will offer a basic model with a range of enhancement options at
additional cost. These options will include larger and more powerful engines, alloy wheels,
leather interiors, more electric and electronic gadgetry and so on. By offering a range of different
versions of the car at different prices the business is able to capture a wider range of customers
and to make more profit from those customers willing to pay for the higher specifications.

3) Premium pricing
Premium pricing is a strategy of setting prices at the top end of the range offered by competitors.
A business using this strategy is not trying to compete on price. Rather, it will concentrate on
competing through the other elements of the marketing mix, such as product quality or features,
ease of access to the product or service (ie availability) or the way the product is promoted.
Establishing a strong brand name to reduce price elasticity enables a business to charge
premium prices. Strong branding also reduces the impact of substitute goods. Coca-Cola is an
example of strong product branding.
Ethical products also tend to be premium priced as they are meeting non-price-based
consumer needs.

Extract 8.3: Premium pricing at MV Agusta


The iconic Italian motorcycle brand MV Agusta has a reputation built upon a racing heritage, stylish Italian
design and high performance. The company opts for a niche marketing strategy that emphasizes this
reputation and is enhanced by celebrity customers such as actors Brad Pitt, Tom Cruise and Angelina Jolie,
and former Formula One world champion Michael Schumacher.
The company has a strategy of maintaining the essence of this brand and focusing on high-quality product
development. Part of this strategy is charging premium prices for its products.

4) Stratified pricing: Price discrimination and market


segmentation
Price discrimination is the practice of charging different prices to different customers for the
same product or service. This practice is common in a number of industries. For it to be
successful it is important that the market for the product and service can be successfully
segmented such that customers are not able to move between the different price brackets. One
example of market segmentation is peak and off-peak travel. Customers are charged different
prices for the same service depending on when they use that service. A business using this
strategy is taking advantage of the fact that those customers who will need to travel during peak
commuting times have little flexibility about when they travel. There is therefore relatively little
elasticity of demand, so prices can be increased. On the other hand, customers travelling for
leisure generally have more flexibility about when they travel. They can be attracted to use buses
or trains during off-peak times by making this cheaper.
A second form of stratified pricing involves producing a range of similar products but with
different features to meet the needs of different customers. An example of this is the different
classes of air travel available. Three passengers may all fly from Paris to New York at the same
time on the same aeroplane. However, each will pay a different price (and receive a different
level of service) by flying economy class, business class or first class. The airline divides the
market for air travel into three different market segments.

Extract 8.4: Stratified pricing wins Apple the US market


The world of mobile phones is dominated by two major players: Apple and Samsung. Apple is known for its
premium pricing strategy and perhaps because of this Samsung has achieved a higher volume of smartphone
sales across the world. However, towards the end of 2012 Apple ended Samsung’s four-year lead in the US
mobile phone market. With the launch of the iPhone 5, Apple also offered lower-priced, old versions of the
iPhone, shipping an estimated 17.7 million phones to the United States in the last three months of 2012. This
stratified pricing strategy provided a range of phones to satisfy the needs of more budget-conscious
consumers. Apple’s share of the market increased from 25 to 34 per cent in one year.

5) Competitive pricing
A competitive pricing strategy will be adopted by a business operating in a market that is close to
perfect competition. These market conditions make it difficult for the firm to charge prices that
are significantly higher or lower than those of competitors. The business will therefore price its
products very close to the price of competitors.

6) Value pricing
Value pricing involves selling a product which is of similar quality to that of competitors at a
lower price. The aim of this strategy is to attract customers by offering better value than
competitors. Although this strategy can mean that the business is potentially not maximizing its
profit from each sale, the aim is to increase sales volume so that overall profitability is increased.
This strategy uses the consumer psychology in the product positioning model discussed earlier in
this chapter.

Extract 8.5: Value pricing at Amazon – a strategy for market monopoly


Amazon, the online retailer, has established a business model that enables it to operate at very low profit
margins – often 2 per cent or less – whilst delivering a high-quality service and cheaper products than the
competition. It is estimated that Amazon now has over one-third of the entire e-commerce market in the United
States. This strong position has enabled Amazon to undercut new market entrants and increase its dominance
of the e-commerce world. In addition, the online retailer has steadily squeezed out high-street bookshops and
DVD retailers.

7) Bundle pricing
Bundle pricing is often used by retailers to entice customers to spend more. Typical bundle
pricing strategies include ‘buy one get one free’ (BOGOF) or ‘three for the price of two’. Such a
strategy decreases the profit margin on each product sold, but increases the volume of sales such
that the retailer can achieve an overall higher level of profit. Usually this is a short-term pricing
strategy.
Another strategy commonly used by retailers, particularly supermarkets, is that of the loss
leader. A loss leader is a product which is priced below cost in order to attract customers into the
store. The rationale behind the loss leader is that it will attract customers who will then buy other
products whilst in the store. This must be a short-term pricing strategy for any one product.
In some cases, prices may be deliberately set very low, even giving away products for free, by
a dominant business in the market in order to cut out competitors. This is known as predatory
pricing, a practice which is illegal under competition law in most legislations.

8) Penetration pricing
Market penetration pricing is the strategy of introducing a new product at a very low price in
order to entice consumers to buy that product. In some cases, this may involve selling at a price
which is lower than the cost of production. (As mentioned above, this is known as a loss leader.)
Once the product is established in the market, the price can be brought back up in line with
competitors’ prices. This is therefore a strategy for increasing market share through sales volume
and will only succeed if there is a relatively high level of price elasticity.
This is a good strategy to adopt when launching a new product into an existing market which
already contains several similar products. The goal of the business launching the new product is
to encourage customers to switch to their product because of its lower price. An example is the
launch of a new washing powder. There are already several different brands of washing powder
on the market. By launching the new washing powder at a low price the business should achieve
two important objectives. Firstly, they will make their product more attractive than that of their
competitors. Secondly, by gaining relatively high levels of sales very quickly they will be able to
establish economies of scale, thus reducing the cost per unit. Once customer habits have been
established in buying the new product, the business can increase the price so that it is in line with
those of competitors.

9) One-off pricing
One-off pricing has been kept as a separate category here because it relates to certain unusual
pricing situations. It is at the bottom end of the pricing scale in Figure 8.8 because the strategy
usually (but not necessarily) involves selling at a very low price. This pricing strategy will be
used by a business that has surplus inventory that could not otherwise be sold. Rather than
setting a price to recover the full cost of the product, the business recognizes that any price which
covers selling costs will put the business in a better financial position than if it kept the otherwise
unusable inventory. This principle is illustrated in Worked Example 8.5.

WORKED EXAMPLE 8.5 One-off pricing


Glaze Co manufactures window units. The company took an order for 50 windows from a customer that subsequently
went bankrupt and never paid for or collected the windows. The windows cost $20,000 to manufacture. Another
customer has placed an order for similar windows. Glaze Co could change the specification of the 50 windows it
already has in stock to meet the needs of a new customer at a further cost of $5,000. It has no other use for the
windows which otherwise would be scrapped. Glaze Co usually adds a mark-up of 30 per cent to the cost of windows
in arriving at a sales price.
What is the minimum price that Glaze Co should accept from the new customer for the 50 windows?

Answer
In this case Glaze Co should ignore the $20,000 it has already spent on manufacturing the windows. If it costs only
$5,000 to adapt the windows to the needs of the new customer, any price above $5,000 will put Glaze Co in a better
financial position than not selling the windows and having to scrap them. The company should therefore be happy to
negotiate any price above $5,000.

Exercises: now attempt Exercise 8.1 on page 322

Target pricing and target costing


Target pricing is an approach to pricing which combines the three elements discussed in this
chapter (cost, competitors and customers). It does so in a manner which recognizes the
importance of cost on profitability, whilst ensuring that the competitive focus on customers and
competitors leads cost considerations rather than being subsumed by them. The technique
originated in Japan in the 1960s, but by the 1980s it was being widely used by a range of
businesses across the world.
Target pricing involves examining the market and customer preferences in order to determine
in advance what the optimum price would be. The business then sets itself the target of
producing the goods or services at a cost which will enable a profit at the target price. This
process typically involves the following five steps:

1. Develop a new product based upon analysis of customer needs and demands.
2. Set a target price that is based upon customers’ perceived value of the product.
3. Set a target profit margin. This will be based upon the required return on investment. (See
Chapter 6 for more details of return on investment.)
4. Derive the target cost of production by subtracting the target profit from the target price.
5. In many cases there will be a cost gap between actual cost per unit and the target cost per
unit. Techniques such as value engineering and kaizen (see below) can then be used to
bring actual costs as close as possible to target costs.

This process is illustrated in Figure 8.9.

Figure 8.9 The target costing process

Value engineering
Value engineering is a customer-focused approach to product design. It involves identifying
those parts of a product or service that add value and equally importantly eliminating those
features which do not add value. The aim of value engineering is to maximize value to the
consumer whilst minimizing costs.
Value engineering focuses on the planning and design stage of a product’s life cycle and
strives to ensure that products are designed in such a way that it is possible to produce them to
target costs. Some features, although desirable, may be much too expensive to manufacture.
Sometimes minor changes at the design stage can create huge efficiency savings once
manufacturing commences. If such matters are not dealt with at the planning stage, it may be
impossible to meet target costs no matter how efficient the production process is.
In practice, activity-based costing will be an important tool at the product design stage, as it
will help in decision making about how product features will translate to manufacturing costs.
By understanding the cost drivers a business will be better able to control its costs.

Kaizen
Kaizen is a Japanese term which refers to a philosophy of continuous improvement in operations.
In practice, it means continually examining the manufacturing processes and business systems of
the organization in order to identify and implement efficiency changes which will reduce costs.
The focus of kaizen is typically on small improvements, rather than large and drastic changes to
production processes. An example of a kaizen improvement can be as simple as moving the
location of a parts trolley by one metre in order to improve the efficiency and speed at which an
operative works.

Exercises: now attempt Exercise 8.2 on page 323

Conclusion
This chapter has examined the factors that go into the pricing decision and has set out a range of
pricing strategies used across different industries. In practice, the pricing decision can be very
difficult. It may not be possible to obtain the information needed to make some of the
evaluations set out in this chapter. Furthermore, many markets are volatile and the factors
underpinning pricing are constantly changing. The pace of change of technology and new
innovation in many products and services makes the life cycle of a product and the competitive
environment extremely uncertain. This means that managers must continually review their
pricing strategies and the costing, competitor and consumer factors which underpin them.

COMPREHENSION QUESTIONS
1. Distinguish between the three main perspectives on pricing and identify the main concerns of each perspective.
2. What are the 3C’s of pricing and how does each impact upon price?
3. Distinguish between full-cost and marginal-cost pricing.
4. Explain the concept of life-cycle costing and how this relates to pricing.
5. If a product is known to have a low price elasticity of demand and a company reduces its price in order to stimulate
demand, would you expect this to lead to increased or decreased overall sales revenues?
6. Publisher Co sells both paper books and e-books. If the company reduces the price of its e-books, should it expect
demand for paper books to increase or decrease?
7. Explain how customers’ expectations can impact upon the price they are willing to pay for a product.
8. What is price skimming and under what circumstances is it an appropriate pricing strategy?

Answers on pages 323–324

Exercises
Answers on pages 325 – 326

Exercise 8.1: Pricing strategy


XTA Co is about to launch a new personal computer called the Eye which the user wears like a pair of glasses.
Interaction with the computer is through eye movement and voice. No competitors are offering such an
advanced product, but two competitors are expected to launch a similar device within the next six months.
Market research has indicated that there is a great deal of customer excitement about this new computer and
there have been several favourable articles in leading computer magazines.

Required:
Set out a pricing strategy for the new product, explaining why you consider it to be the most appropriate.

Exercise 8.2: Target pricing


ACT Motors, a large vehicle manufacturer, is seeking to launch a new family saloon car. The market for such
cars is already well established and highly competitive.

Required:
Explain the main steps ACT Motors should take in developing a target price for the new family saloon car.

Exercise 8.3: Marginal-cost-plus pricing


Printit Co provides rapid printing and design services to the public and small businesses. The business, which
has 12 employees, operates out of a high-street shop which includes office space. The company undertakes
400 to 500 jobs each month. It currently operates a full-cost-plus pricing system, and as each job is unique it
must be priced individually. The accountant, who works part-time, is struggling to keep on top of the task of
pricing each job. She has suggested that the company should move to a marginal-cost-plus pricing system as
this would make pricing much easier.

Required:
Discuss the advantages and disadvantages for Printit Co of changing to a marginal-cost-plus pricing system
from its existing full-cost-plus pricing system.

Answers to comprehension questions


1. The three main perspectives on pricing are the accountant’s perspective, the economist’s perspective
and the marketer’s perspective. The accountant is primarily concerned with the relationship between
revenues and costs and seeks to set a price which will generate the desired level of profit. The economist
is primarily concerned with the competitive environment and the impact which price will have upon the
behaviour of competitors and customers. The marketer is primarily concerned with the way in which a
customer perceives the goods or services provided and sees price as one element that contributes to
that perception.
2. The 3C’s of pricing are: costs, customers and competitors. A business must set the price which covers its
costs, is attractive to customers and enables it to compete with competitors.
3. Full-cost pricing involves calculating the full cost of producing an item for sale. This will include an
apportionment of indirect and overhead costs. Marginal-cost pricing involves calculating only the direct
costs of an item and adding sufficient mark-up to cover indirect and overhead costs. Each method has
advantages and disadvantages and which is the most appropriate will depend upon the circumstances of
the business.
4. Life-cycle costing involves costing a product over its entire life cycle rather than calculating the immediate
cost of producing a unit now. The aim of using life-cycle costing for pricing is to ensure that a product
makes a profit overall. It is more appropriate than traditional techniques for pricing those products which
have a short life cycle and involve high pre-production costs. For such items the pre-production costs
may be far more significant than ongoing production costs and it is important to ensure that these are
incorporated into the price to ensure overall profitability.
5. If the product has a low price elasticity of demand, a change in price will result in a relatively small
change in demand. Therefore, if the price is reduced, demand is unlikely to increase by very much. As a
result, overall revenue will be reduced.
6. Paper books and e-books can be seen as substitute goods. Therefore, if the price of e-books is reduced,
we can expect the demand for paper books to decrease.
7. There are a number of aspects of customer expectations which will impact upon price. For example, if
customers expect supply to decrease or stop in the future, they may be willing to pay higher prices to
obtain a product now. Equally, if customers expect prices to increase in the future, they will be willing to
pay more now.
8. Price skimming is usually used at the launch of a new product. The strategy works best for products
which have previously been unavailable or have new features. The strategy involves launching with a
very high price and then gradually bringing the price down over time. This strategy takes advantage of
the fact that the price elasticity of demand is not the same for all customers: some customers will be
willing to pay a high price to have the product immediately, whereas other customers will not buy the
product until the price falls. By starting with a high price and bringing the price down over time, the
company is able to maximize the profit it earns from the product.

Answers to exercises

Exercise 8.1: Pricing strategy


The most appropriate pricing strategy for the Eye would be price skimming. This would involve launching the
product at a relatively high cost. As there is already a lot of marketing excitement about the product, there is
likely to be a substantial number of customers who will be willing to pay a premium price in order to be among
the first users of this new personal computer.
Once competitors bring similar products onto the market, XTA Co can reduce the price of the Eye to make it
more competitive. This strategy will enable XTA Co to maximize profits whilst there is no competition for the
Eye, but then compete or even undercut competitors once they enter the market.
XTA Co will have first-mover advantage in the market. By the time competitors place their products, XTA Co
will already be established with customers. Furthermore, the company will hopefully have achieved sufficient
volume of sales to bring unit production costs down through economies of scale. This will enable the company
to match competitors’ prices or undercut them while still earning higher levels of profit.

Exercise 8.2: Target pricing


ACT Motors should take the following steps in developing a target price for the new family saloon car:

Step one: the company must develop a new car which meets customer needs in order to attract adequate
sales. Market research should be undertaken in order to establish the features which customers most
desire and preferably identify features not provided by competitors.
Step two: a target price should be set based upon an analysis of customers’ perceived value of the new car.
Market research should be undertaken in order to establish an appropriate price. If the new car has
significant new features not available from competitors, a premium pricing strategy may be chosen by
setting the price higher than that of competitors. Alternatively, a value pricing strategy could be adopted
by offering a product which is comparable with those of competitors at a slightly lower price.
Step three: senior management should set the desired profit margin for the new car. This will be based upon
company policy and will be in line with strategic targets for return on investment or return on sales.
Step four: the target cost of the new car will be derived by subtracting the target profit margin from the target
price.
Step five: projected actual costs should be analysed against target cost and steps taken to reduce actual
costs should they be in excess of the target costs. This will involve techniques such as value
engineering and kaizen.
Exercise 8.3: Marginal-cost-plus pricing
The current pricing system used by Printit Co involves identifying all of the costs involved in each job, including
the relevant proportion of overhead costs, and adding the desired profit margin. A marginal-cost-plus system
would need to identify only those costs incurred as a consequence of the job being undertaken. In this case,
that would involve the costs of staff time and any materials consumed.
For a business like Printit Co, overhead costs will include the property, computer and printing equipment.
These costs are unlikely to vary with the number of jobs undertaken.
09
Investment decisions

OBJECTIVE
All organizations have to make investment decisions through the process of acquiring, replacing or upgrading premises,
equipment or vehicles, hiring new staff, investing in training, or changing systems and procedures. Such decisions must
make strategic sense for the organization, but they must also make good financial sense. It is because of this that managers
at all levels find themselves involved in the financial evaluation of investments, whether it be in preparing proposals or
evaluating options. This chapter therefore aims to provide the understanding necessary for managers to participate in the
financial evaluation of investment opportunities.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Assess the financial impact of an investment using traditional techniques.


Evaluate the strengths and weaknesses of the above techniques.
Demonstrate an awareness of alternative techniques.
Explain the role of financial assessment within the wider strategic assessment of investment decisions.

KEY TOPICS COVERED


Traditional investment appraisal techniques.
The strengths and weaknesses of individual appraisal techniques.
The use and abuse of investment appraisal techniques.
Financial analysis within the overall context of investment planning.
Non-financial factors in investment appraisal.
Alternative strategic approaches to investment appraisal.

MANAGEMENT ISSUES
Managers need to be able to interpret and evaluate the results of financial investment appraisal calculations, rather
than perform the calculations. This chapter therefore focuses on those skills.
The apparent sophistication and precision of financial investment appraisal techniques can mask shortcomings and
behavioural aspects of their application. This chapter therefore includes evaluation of frequently used techniques
within a wider behavioural context.

Introduction
Investment appraisal is the process of deciding which projects or assets to invest in. In this
chapter we will look at how to evaluate investments from a financial point of view and how such
financial analysis fits into the overall investment decision process. Investment appraisal should
always be understood within the wider context of strategic formulation and implementation.
Although such matters are beyond the scope of this textbook, this chapter will look at how non-
financial factors, risk levels and wider strategic concerns can impinge upon the more traditional
concerns of the immediate economic return from an investment.
Investment appraisal, sometimes referred to as capital investment appraisal, is concerned with
organizational decisions about investment in equipment, machinery, buildings or other long-term
assets. This can include a range of types of decision such as replacement of existing assets,
investing in new IT or equipment to reduce operating costs, expansion through purchasing new
buildings or equipment, improving delivery service or staff training. However, the principles
apply equally to investments in shares in companies, whether made by businesses or individuals.
Therefore the techniques looked at in this chapter are equally relevant to both businesses and
individuals who are involved in making investments.
The importance of good investment appraisal lies with the strategic and financial importance
of the investments made. Investments will shape the future of the organization. Such investments
often involve large resources. Wrong decisions can be costly and difficult to reverse and will
have a direct impact on the organization’s ability to meet its strategic objectives.
From a financial point of view, an investment involves making a cash outlay with the aim of
receiving future cash flows in return. At a basic level, assessing the financial viability of an
investment involves simply comparing costs with benefits, and ensuring that the benefits
outweigh the costs. However, in practice this can prove difficult, as identifying and measuring
the costs and benefits from an investment can be a complex task. This chapter will address some
of the problems of investment cost–benefit analysis and look at how techniques have developed
in order to meet these problems.
In order to help make investment decisions, a common method of appraisal is required; one
which can be applied equally to a whole spectrum of investment situations and which will enable
the decision-maker to assess individual investments and compare alternative investment
opportunities.

Investment appraisal – the basics


In this section we will consider the important questions which should be asked when an
investment is being evaluated. Imagine that you are offered an investment opportunity as
follows: ‘Invest $1,000 with me now and I promise that I will make you rich.’ Before you hand
over your money, what is the information that you need in order to evaluate this investment
opportunity effectively?
Firstly, you would want some clarification on how ‘rich’ the investment would make you. In
other words, you want some information on the return that the investment offers.
Secondly, you would also want to know when you will become rich. You therefore need more
information on the timing of returns from the investment. There would be a significant difference
between receiving a return immediately and receiving the same return in 10 years’ time. This is
referred to as the ‘time value of money’. This principle recognizes that the sooner you receive
a return, the more valuable that return is to you.
Thirdly, you would also want some more information on the risks involved in your
investment. If you are going to invest $1,000, what is the risk that the promised return never
materializes (or that you lose the whole $1,000)?
Therefore, we can say that in evaluating a potential investment we need information about
three key issues: risk, return, and timing.
Having established these basic principles we can now look at the techniques most frequently
used in practice by businesses for investment appraisal, and examine how these techniques
address the questions raised above.

Traditional evaluation techniques


There are three techniques which are in common usage in evaluating capital investment
decisions:

payback period (PP);


accounting rate of return (ARR);
discounted cash flow (DCF).

The last of these techniques, discounted cash flow, can be divided into two different methods of
application: net present value (NPV) and internal rate of return (IRR).
In order to examine each of these techniques we will look at how they are applied to a simple
investment opportunity, as set out in Worked Example 9.1.

WORKED EXAMPLE 9.1


Soundzgud Co is a manufacturer of hifi equipment. The company is currently considering the launch of a new
amplifier called the ‘Window Rattler’. Research indicates that the company can expect sales of its new amplifier as
shown in Table 9.1.
Table 9.1

Year 1 2 3 4 5 6
$’000 $’000 $’000 $’000 $’000 $’000
Revenue 600 800 800 700 500 400

The machinery needed to produce the ‘Window Rattler’ will cost $1,200,000. However, at the end of six years this
machinery can still be used for other products and is expected to then have a value of $150,000. The machinery will
be purchased as soon as the decision to manufacture has been approved by the board of directors.
The company accountant has forecast the production costs for the ‘Window Rattler’ (Table 9.2).

Table 9.2

Year 1 2 3 4 5 6
$’000 $’000 $’000 $’000 $’000 $’000
Production costs 320 410 410 350 250 200

The production costs exclude depreciation, which the company normally charges on a straight-line basis. The
company uses a discount rate of 15 per cent to evaluate new investments.

Expert view 9.1: Common notation conventions


Because of the importance of timing, any cash flows or profits in relation to an investment must be allocated to
a particular accounting period. With the use of computer spreadsheets, an accurate analysis of timing can be
made to the nearest month or even week or day. However, as all cash flows used in investment appraisal are
forecasts, and therefore by their very nature estimated, it is often not possible to identify when they will occur to
such a high level of accuracy. Therefore, the most common convention for investments which will run over
several years is to identify cash flows by year and to assume that they will arise at the end of the year.
Obviously, this may not be the actual pattern of cash flows.

A significant difference between profit and cash flow is that profit is recorded using the accruals
principle. That is to say, sales and purchases are recorded when the transaction takes place rather
than when any underlying payment is made. This means that, in practice, profit may be allocated
to a different period from the underlying cash flow. However, for simplicity, in this example it
will be assumed that cash flows occur in the same period in which accounting costs or revenues
arise.
Another significant difference between profit and cash flow is depreciation. This is an
accounting adjustment rather than a cash payment.
A common convention in recording the timing of cash flows or profits is to use the term t0 to
denote an item which occurs immediately, t1 to denote an item which occurs in one year’s time,
t2 for two years’ time and so on. Hence, the cash flows forecast for the ‘Window Rattler’ can be
recorded as shown in Table 9.3.

Table 9.3
Timing Cash flow $’000
t0 (Immediately) Cost of machine (1,200)
t1 (1 year’s time) Net cash inflow (600–320) 280
t2 (2 years’ time) Net cash inflow (800–410) 390
t3 (3 years’ time) Net cash inflow (800–410) 390
t4 (4 years’ time) Net cash inflow (700–350) 350
t5 (5 years’ time) Net cash inflow (500–250) 250
t6 (6 years’ time) Net cash inflow (400–200) 200
t6 (6 years’ time) Residual value of machine 150

We can now go on to examine how each of the investment appraisal techniques deals with
this information.

Payback period (PP)


The payback period is the length of time it takes for an initial investment to be repaid out of the
net cash inflow from a project. The easiest way to calculate this is to establish the cumulative net
cash-flow position at the end of each year of the investment. Applied to the example of the
‘Window Rattler’ the payback period can be calculated as shown in Table 9.4.

Table 9.4

Timing Cash flow $’000 Cumulative $’000


Immediately (t0) Cost of investment (1,200) (1,200)
1 year’s time (t1) Net cash inflow 280 (920)
2 years’ time (t2) Net cash inflow 390 (530)
3 years’ time (t3) Net cash inflow 390 (140)
4 years’ time (t4) Net cash inflow 350 210

There is no need to take this calculation further than four years as we can see that the initial cost
of the investment is paid back sometime between three and four years. At the end of three years
there is still another $140,000 to pay back, but after four years the initial cost plus a further
$210,000 has been paid back. If we stick with the assumption that all cash flows arise at the end
of the year, we would say that this investment has a payback period of four years. However, if
we assume that the cash inflows arise evenly throughout the year, we could calculate the payback
period to a fraction of year as follows:
Simply knowing that this investment would take 3.4 years to pay back does not in itself tell us
whether this is a good investment. An evaluation has to be made as to whether this is an
acceptable payback period. However, if two or more alternative investments are being compared,
they can be ranked in terms of their payback period, with the shortest being the most attractive.

Extract 9.1: Payback period in use – Standard Life and Next


In practice, businesses which use this technique have predetermined payback periods. These tend to differ
according to type of investment. For example, an investment in IT equipment may be required to pay back
within two years. On the other hand, investment in a new building may be required pay back within 20 years:

Standard Life, a pensions and life assurance company, set itself a payback period of five years for any
new products launched.
Next, a UK-based retailer, uses a payback period of 18 months when investing in a new clothing store.
However, when the business opened its first Home and Garden store in 2011, it set a payback period of
25 months.

Evaluation of PP
Having examined how the payback period works as an investment technique, we can evaluate
how well it addresses the three questions we set at the beginning of this chapter. The payback
period, by its very nature, tells us something directly about timing. In doing so, it also tells us
indirectly about an important aspect of risk, because a shorter payback period can be equated
with a lower level of risk. However, the technique tells us very little about return other than the
fact that the initial outlay is recovered. Payback period can therefore be seen as more of a risk
appraisal tool than a measure of return.
In fact, in terms of addressing return, the technique has some major shortcomings. Firstly, it
ignores cash flows outside of the payback period and in doing so ignores total return. Consider
Worked Example 9.2 which compares two potential investments.

WORKED EXAMPLE 9.2

Table 9.5

Timing Cash flow Investment A: $’000 Investment B: $’000


Immediately (t0) Cost of investment (600) (600)
1 year’s time (t1) Net cash inflow 200 200
2 years’ time (t2) Net cash inflow 200 200
3 years’ time (t3) Net cash inflow 200 200
4 years’ time (t4) Net cash inflow 10 400
5 years’ time (t5) Net cash inflow 5 600
6 years’ time (t6) Net cash inflow 5 800

Both Investment A and Investment B have identical payback periods of three years and are therefore equally
attractive according to the payback period technique. However, in years 4 to 6 the cash returns of Investment A fall
substantially. In comparison, the cash returns of Investment B continue to rise. Looking at this full picture, Investment
B is clearly a better alternative, as the same initial investment offers a much greater overall return. But this fact is not
revealed by the payback period method of analysis. To take this example a step further, consider the implication of
Investment A offering a $400,000 return in year 1. This would make the payback period two years. Investment A
would therefore be favoured using the payback period method of evaluation, even though Investment B offers a much
higher overall return. An issue to be aware of when using this technique, therefore, is that it favours short-term returns
rather than highest overall returns.

A second shortcoming is that the technique ignores the timing of cash flows within a payback
period. Consider Worked Example 9.3, again of two investments with identical payback periods.

WORKED EXAMPLE 9.3

Table 9.6

Timing Cash flow Investment A: $’000 Investment B: $’000


Immediately (t0) Cost of investment (600) (600)
1 year’s time (t1) Net cash inflow 500 50
2 years’ time (t2) Net cash inflow 50 50
3 years’ time (t3) Net cash inflow 50 500

Both investments take exactly three years to pay back the initial cost. However, Investment A has repaid the bulk of
this initial cost by the end of the first year. There is therefore a significant difference between the two investments in
that we can say Investment A is less risky. Were the investment to stop suddenly for some reason at the end of year
2, the business making Investment A would have already recovered a substantial part of the initial cost, whereas the
business making Investment B would have lost most of this money.

Despite the shortcomings, payback period is an extremely popular investment appraisal


technique in practice. Several surveys of businesses conducted over the past 40 years have all
shown that, despite the rise in popularity of more sophisticated techniques, payback period
remains the most widely used means of investment appraisal, being used by around 80 per cent
of businesses. Its strength lies in its simplicity – it is not difficult to calculate or to understand.
Most importantly, it appeals to a basic human level of psychology. Anyone having made a
substantial investment understands that there is an initial period of anxiety concerning the risk of
that investment. The point at which the investment has repaid its initial outlay is an important
psychological landmark as it means that the initial capital investment has not been lost. It is
because of this that payback period becomes an extremely important technique when a company
has liquidity constraints or severe limitations on the availability of financing. The technique is
also most useful for projects which are known to have a short life and require a quick repayment
of investment.

Extract 9.2: Disney buys Star Wars


In 2012 Disney bought Lucasfilm, the company behind the Star Wars films, for US $4bn. How does this
purchase price stack up in terms of return on that investment? The main value comes from the potential
earning of future films. Disney has announced that it plans to release three new Star Wars films, one every
three years, starting in 2015. The last three Star Wars films generated around US $1.5bn each at the box
office. Such films usually cost around US $0.5bn to make and market, so that would generate Disney net cash
flows of US $3bn. In addition, Lucasfilm has ongoing revenue of around US $0.9 billion per year from existing
films, video games and related consumer products. There is an obvious risk in terms of the success of future
films, but the last three Star Wars films were a box office success despite not being well received by many
fans. These cash-flow forecasts would suggest a payback period as short as five years. Not bad for a brand
that has been financially successful for the past 35 years and looks set to continue long into the future.

Accounting rate of return (ARR)


The term ‘accounting’ in accounting rate of return refers to the basis of calculation of this
technique, which is accounting profits. Unlike the other techniques examined in this chapter
which all use cash flow, ARR is calculated on the accruals basis, ie using profit.
The ARR is sometimes referred to by other names. In the United States it is more widely
called return on investment (ROI). Other names include average rate of return, book rate of
return or unadjusted rate of return. It is also sometimes called return on capital employed
(ROCE). However, as will be seen below, this name is more appropriately applied to divisional
or overall company performance, rather than the tool for investment appraisal. This multitude of
names reflects both the wide usage of the technique and the fact that there is a range of
definitions as to its calculation.
Although there are different ways of calculating ARR, which will be discussed below, we will
concentrate on the most commonly applied calculation. This takes the ‘return’ on an investment
as being the average accounting profit after depreciation and expressing this as a percentage of
the average investment over the life of the project:

Applying this formula to our investment scenario in Worked Example 9.1, the average annual
profit can be calculated by dividing the total profit for the investment by the number of years
over which this profit is earned.
The total profit before deducting depreciation is $1,860,000 ($280,000 + $390,000 +
$390,000 + $350,000 + $250,000 + $200,000).
The total depreciation can be calculated as the initial cost less the residual value of the
investment = $1,200,000 – $150,000 = $1,050,000.
Because the company charges depreciation on a straight-line basis, the average investment will
be the value halfway between the initial cost and the residual value:

Once calculated, this ARR of 20 per cent can be compared with a predetermined minimum
acceptable return for the company. For example, if Soundzgud has a predetermined requirement
of a return on investments of 15 per cent, the above return of 20 per cent would be acceptable.

Evaluation of ARR
Unlike payback period (and the discounted cash-flow techniques we will look at later in the
chapter), the ARR evaluates investments based upon profits. The other techniques use cash
flows. There are both advantages and disadvantages to this use of profit.
The main advantage of ARR is that it calculates the performance of an investment in terms of
profit returns, which is in line with the most often reported figure for overall company
performance. ARR provides a measure which is directly comparable with ROCE, which is the
most common measurement of the financial performance of a business as a whole.
ROCE is measured as:

It can therefore be seen that there is direct comparison between the formula for ARR and that for
ROCE. They are both measuring the same thing – the level of profit in relation to the capital
invested in order to earn that profit: ARR measures this at the level of the individual investment;
ROCE measures it at the divisional or company level.
It makes sense to use a measure for evaluating new investments which reflects the way the
performance of the whole business will be evaluated. By comparing the ARR of a new
investment with the existing ROCE of the whole business, managers are able to assess the
potential impact of that new investment on the financial performance of the business as a whole.
If a new investment has an ARR which is lower than the existing ROCE, that investment will
reduce the future ROCE of the business. On the other hand, a new investment with an ARR
greater than the existing ROCE will increase that ROCE in the future.
Despite the advantages outlined above, ARR suffers from some major deficiencies which
have caused the technique to be criticized by many academic commentators. However, it is still
widely used in practice, so you need to be aware of these deficiencies and their implications:

ARR uses the accruals concept, that is to say, it is calculated from profit rather than the cash
flow. Profit is far more judgemental and therefore easily manipulated than cash flow. For
example, profits from an investment will vary with different methods of stock valuation and
depreciation calculation. This means that it is more difficult to obtain an objective measure
of performance. If ARR is being used to make comparisons between different organizations,
problems will arise in comparing the figures.
ARR is a relative measure: it measures profit returns in relation to the amount invested. This
makes it possible to compare the profitability of investments of different sizes, but it means
that if ARR is used to choose between two or more projects, a small project may have a
higher ARR than a larger project whilst giving a much smaller absolute profit.
There is no universally accepted basis of calculating the figures used for either the profit
from an investment or the capital invested to earn that profit. This means that, in practice,
erroneous comparisons may be made between investments because one is not comparing
like with like. One major problem is that there is disagreement over whether the ARR
calculation should use average capital invested (as shown in the example above) or initial
investment. The argument for using the initial investment is that this is the cost of the
investment and therefore reflects the return required to recover that cost. The more widely
accepted average capital invested method argues that the ARR calculation should take into
account the fact that the initial cost is written off over the life of the investment, and that an
average return therefore needs to reflect the average value of the capital investment. Neither
method is wrong; they simply reflect different principles.
The technique ignores the timing of profits, as the averaging used in the calculation of ARR
removes all information about timing of receipts and payments. Consider Worked Example
9.4 in which three different investments each lasts four years:

WORKED EXAMPLE 9.4

Table 9.7

Profits ($’000) Investment A Investment B Investment C


Year 1 850 50 250
Year 2 50 50 250
Year 3 50 50 250
Year 4 50 850 250

All three investments offer an average annual profit of $250,000. However, there are significant differences in both the
evenness and the timing of this profit. This information, which may be important in an investment decision, is lost in
the ARR calculation through the process of averaging. ARR is therefore blind to the time value of money.
Another problem arises because of this averaging process. It renders the technique
unsuitable for comparing investments of different lengths. If an investment offers long-term
but diminishing profits, then as each year of a lower profit is added to the calculation the
average profit will be pulled downwards. Consider Worked Example 9.5 in which two
investments have the same initial cost of $1 million with no residual value. However,
Investment A has a life of only three years whereas Investment B has a life of six years:

WORKED EXAMPLE 9.5

Table 9.8

Profits ($’000) Investment A Investment B


Year 1 300 300
Year 2 300 300
Year 3 300 300
Year 4 – 200
Year 5 – 150
Year 6 – 100

For the first three years both investments offer the same level of profits. However, Investment B continues to deliver
profit for a further three years, albeit at a lower level. Investment B offers a total profit return of $1,350,000 from an
initial investment of $1 million, whereas investment A offers a total profit return of only $900,000 from the same initial
investment.
However, when we calculate the ARR of each investment, Investment A appears to be the more attractive simply
because the average annual profit is higher:

Discounted cash flows and the time value of money


The final two investment appraisal techniques we will examine in this section both involve the
use of discounted cash flows. Discounting a cash flow involves making adjustments for the
time value of money, that is to say, to reflect the fact that when you receive money has an impact
upon its value to you.
Here is a simple example to illustrate this principle: Would you prefer to receive $100 now or
$100 in two years’ time? I would imagine that you would answer that you would prefer to
receive the $100 now, for a number of reasons:

There is always the risk that you will not actually receive the $100 in two years’ time.
The effect of inflation means that you can buy more with your $100 now than you will be
able to in two years’ time. Hence it is more valuable to you now.
You could take the $100 now and invest it for two years, after which you would have more
than $100.

We will look at risk and inflation later in this chapter, but at this point we can examine the issue
of earning interest in more detail. In the example above, deciding between $100 now or $100 in
two years’ time was relatively easy. However, what if I offered you a choice of $100 now or
$108 in two years’ time? You need some means of comparing the two amounts to decide which
is the more attractive. One way in which you could do this is to ask: ‘What would $100 received
now be worth if I invested it for two years?’

Compounding and discounting


Let us assume that you can invest your $100 received now at an interest rate of 5 per cent. At the
end of one year you will have $105 ($100 × 1.05). At this point, the principle of compound
interest starts to apply, because in the second year you will earn interest not only upon your
original $100 but also upon the $5 interest you earned in the first year. Hence, at the end of two
years you will have $110.25 (($100 × 1.05) × 1.05).
Because the formula calculating this is $100 × 1.05 × 1.05 = $100 × 1.052, we can generalize
this as:

This technique gives you a way of comparing the $100 offered now with the $108 offered in two
years’ time. You can calculate that if you take the $100 now and invest it for two years at 5 per
cent, you will end up with $110.25. You would therefore be better off taking the $100 now rather
than the $108 in two years’ time.
This technique works well for a simple investment scenario such as that given above.
However, it is not so useful if the investment is a little more complex.
What if you were offered the alternative of $100 now or $55 in one year’s time and a further
$55 in two years’ time? In this case you cannot use the compounding technique shown above,
because there is no single time in the future to which you can compound the $100 received now.
We therefore need to modify the technique.
The one point in time which we can always use consistently is now (the present moment). So
rather than asking the question ‘what will $100 received now be worth at some point in the
future?’, we ask ‘what would the money received in the future be worth if received now?’ By
asking the question this way we can compare monies received at several different points of time
in the future.
Let us go back to the choice between $100 received now and $108 received in two years’
time. Rather than compounding the $100 forward, we discount the $108 back to today’s date
by asking the question: ‘What would I need to receive now to have the equivalent of $108 in two
years’ time?’ Another way of looking at this is to ask: ‘If I can earn interest at 5 per cent per
year, how much do I need to invest now in order to end up with $108 after two years?’ We can
calculate this by using the formula which we established above:

A = P × (1 + r)n

We used this formula before to calculate the value of A (the amount received in the future) when
we already have the value of P (the amount invested now). We simply need to rearrange the
formula to start with A, the amount receivable in the future ($108), and calculate the value of P:

This means that if we wanted to receive $108 in two years’ time we would have to invest $97.96
now. In terms of the time value of money, receiving $108 in two years’ time is the equivalent of
receiving $97.96 now. We can compare the two alternatives: receive $100 now or receive an
amount in two years’ time which is the equivalent of receiving $97.96 now. We end up with the
same decision as we did before – it is better to take the $100 now.
What of the second scenario in which you were offered the alternative of $100 now or $55
received at the end of one year and a further $55 received at the end of two years? The
calculation is a little more complex, but we can evaluate these alternatives using this new
discounting technique. We do this by taking each of the future amounts and discounting them
back to their present value and then adding them together as follows:

Therefore, receiving $55 in one year’s time plus $55 in two years’ time is the equivalent of
receiving $52.38 + $49.89 = $102.27 now. We are now able to evaluate this choice. In this case
we would be better off taking the two future amounts as their present value is more than the $100
offered now. (In fact, we can measure precisely that we would be $2.27 better off in today’s
terms by taking the future amounts: $102.27 – $100 = $2.27).

Discount tables
The fraction by which we multiply the future cash flow A in order to calculate its present value P
is known as the discount factor. This, as was shown above, is calculated using the following
formula:

Because we use the same formula every time to calculate a discount factor, a standard set of
values can be set out in a table which provides the discount factor for common discount rates and
time periods. A discount factor table is available in Appendix D.

Net present value (NPV)


The NPV technique uses the principle of discounting cash flows explained above. The NPV of
an investment is the sum of the present values of all cash flows which arise as a result of
undertaking that investment.
The present value (P) of a single sum, A, receivable in n years’ time, given an interest rate
(discount rate) of r, is given by:

Let us now see how this technique works when applied to the Soundzgud Co investment in
Worked Example 9.1. The discount rate we need to apply to the cash flow from the project is 15
per cent. The relevant discount factors have been extracted from Appendix D. We can present the
NPV calculation as shown in Table 9.9.

Table 9.9

Cash flow Discount factor Present value


$ (15%) $
t0 (Immediately) (1,200,000) 1.000 (1,200,000)
t1 (1 year’s time) 280,000 0.870 243,600
t2 (2 years’ time) 390,000 0.756 294,840
t3 (3 years’ time) 390,000 0.658 256,620
t4 (4 years’ time) 350,000 0.572 200,200
t5 (5 years’ time) 250,000 0.497 124,250
t6 (6 years’ time) 200,000 0.432 86,400
t6 (6 years’ time) 150,000 0.432 64,800
Net present value $70,710
From this calculation it can be seen that the NPV of this investment is $70,710. This can be
interpreted as meaning that the company will be $70,710 better off, in today’s terms, by
undertaking this investment.
There is therefore a simple rule for interpreting NPV calculations: If the NPV is positive, the
company will increase its wealth by undertaking the investment, which is therefore financially
viable; if the NPV is negative, the company would decrease its wealth by undertaking investment
and the investment should therefore be rejected.
The NPV technique can be seen as a more sophisticated means of investment appraisal than
the payback period and the ARR. Unlike the payback period, NPV takes account of the entire
cash flow of the project; unlike ARR, NPV takes account of the timing of earnings from an
investment. The further into the future a cash flow arises, the more it is discounted. This reflects
the increased risk and uncertainty of cash flows as they lie further into the future.
Not only is NPV more sophisticated in the way that it takes account of timing, risk, and
return, but it is also a technique which is capable of incorporating more sophisticated and subtle
analysis of investments. We will see this as we look at more complex investment scenarios later
in the chapter.

Internal rate of return (IRR)


Internal rate of return is a second discounted cash-flow technique which works on the same
mathematical principles as NPV, but uses discounting to give an answer in a slightly different
format.
In the NPV calculation above, we discounted the cash flow from the Soundzgud investment at
15 per cent to produce a positive NPV of $70,710. The implication of this positive NPV is that
the actual return of the investment is greater than 15 per cent. The discount rate which we apply
in calculating the NPV represents the minimum acceptable return. A positive NPV means that
this minimum return has been exceeded and this is why investments with a positive NPV should
be accepted.
Another way of using the DCF technique would be to calculate the actual discounted cash-
flow return of the investment and compare that to our minimum acceptable return of 15 per cent.
This is known as the internal rate of return, as it is the rate of return ‘internal’ to, ie within, the
project.
In practice, the IRR will be the discount rate which gives an NPV of zero. With a
computerized spreadsheet, calculating this discount rate is relatively straightforward. However, if
a computer is not available, the IRR can be estimated as demonstrated below.
Returning to the Soundzgud investment in Worked Example 9.1, the NPV at 15 per cent was
a positive value of $70,710. This means that the IRR of the investment must be greater than 15
per cent. We can therefore choose a discount rate greater than 15 per cent and recalculate the
NPV to see if we are closer to the IRR. (Remember, the exact IRR would give an NPV of zero.)
Let us discount the project again using a discount rate of 20 per cent (Table 9.10).

Table 9.10
Timing Cash flow Discount factor Present value
$ (20%) $
t0 (Immediately) (1,200,000) 1.000 (1,200,000)
t1 (1 year’s time) 280,000 0.833 233,240
t2 (2 years’ time) 390,000 0.694 270,660
t3 (3 years’ time) 390,000 0.579 225,810
t4 (4 years’ time) 350,000 0.482 168,700
t5 (5 years’ time) 250,000 0.402 100,500
t6 (6 years’ time) 200,000 0.335 67,000
t6 (6 years’ time) 150,000 0.335 50,250
Net present value $(83,840)

This time we get a negative NPV of –$83,840. This means that the IRR must lie somewhere
between 15 and 20 per cent. We can establish the IRR more accurately using a mathematical
technique called linear interpolation. This is done by applying the formula:

So in this case we can say that the investment offers an internal rate of return of 17.3 per cent.
We can then compare this to an acceptable minimum level of return in the same way as we did
with the ARR technique.

Extract 9.3: EDF – IRR in the energy industry


Investment in energy infrastructure in the UK involves a delicate partnership between the government and the
private sector. Adequate returns are critical to attracting suitable private-sector investment. The government is
seeking to increase investment in low-carbon energy infrastructure, but this can represent a high risk to
investors who consequently require high returns. Investors in gas-fired power plants typically expect an IRR of
around 10 per cent, whereas investors in wind generation expect an IRR of 10 to 13 per cent.
The government can reduce the returns demanded by reducing the risk. In 2013 the UK government struck
a deal with EDF Energy, the French energy company, to subsidize the building of a new nuclear power plant at
Hinkley Point in Somerset. The plant has an estimated build cost of £16bn, which is too high to represent a
viable investment. The government is therefore offering various incentives, including guaranteeing EDF’s
income in order to create an IRR of 10 per cent from the plant. Industry experts estimate that this will add up to
a subsidy of £7 per household per year. The complexity of this deal underscores the importance of IRR for
investors. Without an adequate return on the investment, no new power plants would be built and the UK would
risk facing power shortages.
Evaluation of IRR
The IRR is used in the same way as ARR, by comparing with a predetermined acceptable value.
However, the IRR uses discounted cash flows rather than average profits, and therefore has all
the advantages we identified when looking at the NPV technique: it takes account of risk, timing
and returns.
However, one major drawback of the IRR technique is that it will not work with investments
which have what are known as ‘non-conventional’ cash flows. A conventional investment cash
flow is one which involves an initial cash outflow followed by a series of cash inflows. With
some investments, net cash flows can flow both inward and outward throughout the life of the
project. For example, investment in a nuclear power plant may involve initial cash outflows
followed by many years of cash inflows and then substantial cash outflows as the power plant is
decommissioned at the end of its life. The IRR technique cannot cope with cash flows such as
this, as mathematically it will produce more than one value. It is possible to modify the IRR
technique to work around this problem, but such calculations are beyond the scope of an
introductory text such as this.
Although DCF techniques are more sophisticated in integrating the factors of risk, timing and
returns, they do have some drawbacks which have led to criticism. The process of discounting
cash flows inevitably leads to a favouring of investments which offer returns in the shorter term.
It has been suggested that this leads to short-termism and a reluctance to make investments
which offer strategic benefits which may be more long-term and more difficult to quantify. Also,
the techniques have been criticized as being inadequate for evaluating new technology
investments as they are unable to evaluate non-quantifiable issues. We will look at some
methods for addressing these problems later in the chapter.

Exercises: now attempt Exercise 9.1 on page 363

Incorporating real-world complexities into investment


appraisal
So far we have examined the main investment appraisal techniques using a relatively
straightforward example. However, in the real world, investment appraisal decisions will be
more complex and need to take into account issues such as inflation and the impact of taxation.
In this section we will examine some of these real-world complexities and look at how they can
be incorporated into our calculations.

Establishing an appropriate discount rate


The discount rate used in an NPV calculation represents the minimum acceptable rate of return
for the investor. In practice, a wide variety of methods can be used to determine appropriate
discount rates. Most organizations establish a discount rate based upon either the weighted
average cost of capital (WACC) or the capital asset pricing model (CAPM). Explanation of these
techniques is beyond the scope of this textbook. However, whatever method is used in practice, it
is based upon certain principles. The rate of return from an investment should be sufficient to
cover the cost of financing that investment and should incorporate an assessment of the risk. In
simple terms, the more risky an investment is perceived to be, the higher the discount rate used
to evaluate it. Hence one important way in which risk assessment is incorporated into investment
appraisal is through adjustment of the discount rate used in NPV calculations.

Deciding what to count: relevant cash flows


One of the most common mistakes in investment appraisal in practice involves using the wrong
figures. This can mean including costs which should not be counted, or omitting costs or
revenues which should be included. These mistakes can lead to inappropriate investment
decisions, with disastrous consequences for future business operations.
An important principle in decision making, therefore, is that only those costs or revenues
which are affected by the decision should be considered. A frequent error is to include costs
which will be incurred anyway, even if the investment were not made. These may include, for
example, the costs of employees working on a project, but who would be paid anyway,
regardless of whether or not the project goes ahead. In a similar way, factory costs,
administration and head office costs which will not change as a result of the project should not
be included.
Likewise, costs already incurred, even if they relate to the investment, such as market research
or product development already completed, should not be included in an investment appraisal
calculation. This is because these costs, having already been incurred, cannot change as a result
of making the investment.
Therefore, as a simple rule of thumb, you should include in investment appraisal calculations
only those costs or revenues which will change as a result of undertaking the investment (see
Worked Example 9.6 below).

Opportunity costs
Opportunity costs are a category of costs used only in decision making. They can be extremely
important in assessing the true financial impact of an investment. An opportunity cost can be
defined as the cost of an alternative that must be forgone in order to pursue a certain action.
Opportunity costs arise from the recognition that committing resources to one project means that
they cannot be used on other projects. This may mean that there is a cost to the business in terms
of the lost ‘opportunity’ of using that resource elsewhere. For example, using a production
facility to make one product means that that facility cannot be used to make a different product,
even though the alternative may be extremely profitable. This example may be obvious, but
some opportunity costs are less obvious. For example, opening a new store may draw business
away from existing stores in the same area, reducing their profits.
By incorporating opportunity costs into investment appraisal calculations, a business is able to
quantify the loss of other potentially profitable opportunities and thereby ensure that the most
profitable course of action is taken. To ensure that this is the case, opportunity cost is always
measured in terms of lost contribution from the use of a particular resource.

Expert view 9.2: Contribution


Contribution = Revenue – Variable (direct) costs

WORKED EXAMPLE 9.6 Relevant costs and opportunity costs


Zebra Co is a paint manufacturer that is planning to launch a new range of fast-drying paints. The directors of Zebra
Co have asked you to help evaluate the financial viability of the new paint range. Which of the following amounts
would you include in an NPV calculation?

A A proportion of head office administrative costs calculated at $18,000 per year.


B Depreciation on the new paint manufacturing plant purchased for the project.
C Salaries of 10 new workers hired to operate the new plant.
D Financing costs of $25,000 per year on the loan to purchase the new plant in (B) above.
E The salary of $35,000 per year of the manager running the new plant. This manager ran one of the other
manufacturing plants and has been transferred to this job because of his experience. In his absence, the other
plant is being run by a new manager hired as a replacement for the duration of this project. The new manager
has a salary of $28,000 per year.

Answers

A Head office costs will be incurred regardless of whether or not the new range of paints is launched. These costs
should therefore not be included in an NPV calculation as they are not relevant.
B Depreciation is not a cash flow. This cost should therefore not be included.
C The salaries of new workers are a relevant cost and should therefore be included in the NPV calculation.
D Financing costs should not be included in the cash flows to be discounted. The cost of financing the investment
will be incorporated into the discount rate used to evaluate the investment.
E This is an example of an opportunity cost. The additional cost to the business of using the experienced manager
on this project is the $28,000 it costs to replace him in his old job. The amount which should be included in the
NPV calculation is therefore $28,000 per year, rather than $35,000 per year.

Taxation
Tax payments can have a significant impact on investment cash flows and should therefore be
taken into account. There are four aspects of taxation which are relevant to investment appraisal:

1. Income tax paid on profits from an investment will represent a cash outflow for a business
and therefore needs to be incorporated into the NPV calculation. Because of the importance
of timing, when tax is paid can make a difference.
2. Interest on any debt financing is an allowable expense against income tax. The method of
financing an investment will therefore impact upon tax cash flows.
3. Any tax losses may give rise to tax relief on other profits which can reduce the amount of
taxation payable.
4. Capital allowances are the income tax equivalent of depreciation. They provide tax relief on
the investment in assets which will reduce the cash outflows for income tax. For projects
involving large investments in buildings, plant or equipment, capital allowances can be
significant and can have a major impact upon the cash flows of the investment.

Because these aspects of taxation can have a significant impact upon the cash flows from an
investment, it is important to ensure that taxation is always considered and incorporated into
investment appraisal calculations.

Inflation
Inflation decreases the purchasing power of future cash inflows, making them worth less.
Therefore inflation can create distortions when attempting to assess the time value of money and
in calculating returns from investments. For example, a rate of return of 12 per cent could be
reduced to 8 per cent in real terms after taking inflation into account. It is therefore important to
incorporate the impact of inflation into the return on investments and to be clear whether a
quoted rate of return includes or excludes inflation.
To avoid this confusion, different terms are used. If an interest rate includes the effect of
inflation, it is referred to as the ‘money rate’. Just to add the potential for confusion, this is
sometimes also known as the ‘nominal’ or ‘market’ rate. On the other hand, ‘real’ interest rates
are stated after adjusting to remove inflation. The adjustment is done as follows:

(1 + real rate of interest) × (1 + rate of inflation) = (1 + money rate of interest)

Note that the real rate is multiplied by inflation, not added, to give the money rate.

WORKED EXAMPLE 9.7 Adjusting for inflation


a The real rate of return which a business requires from its investments is 12 per cent and inflation is currently 4 per
cent. What is the money rate of return which the business needs to apply for investment appraisal?

Answer: The money rate of return = 16.5 per cent (1.12 × 1.04 = 1.165).

b A business experiences a money rate of return of 18 per cent on an investment. Over that same period inflation
was 5 per cent. What is the real rate of return on the investment?

Answer: The real rate of return = 12.4 per cent (1.18 ÷ 1.05 = 1.124).

Dealing with inflation in investment appraisal calculations


There are two ways in which the impact of inflation can be incorporated into investment
appraisal calculations:

Money method: the first method is to use ‘money’ cash flows (ie those cash flows which
include inflation) and to discount these using the ‘money’ discount rate (ie the discount rate
which incorporates the effect of inflation).
Real method: the second method is to use cash flows which exclude the impact of inflation
and to apply the ‘real’ discount rate.

Which of these techniques is used in practice will depend upon how the cash-flow forecast for an
investment has been compiled. If the cash-flow forecast is in ‘today’s terms’, without
consideration of the impact of inflation, it is easier to use the real method. On the other hand, if
cash-flow forecasts have been compiled looking at actual amounts payable or receivable in the
future (and therefore incorporating inflation), the money method should be used.

WORKED EXAMPLE 9.8 Dealing with inflation


A company is considering investing $100,000 in a project which will give returns of $50,000 in current terms for three
years. The money rate of return required by the company is 14 per cent and inflation is currently 4.6 per cent. What is
the NPV?

Money method

Table 9.11

Year Real cash flow Money cash flow Money discount factor Present value
$ (inflated by 4.6%) @14% $
0 (100,000) (100,000) 1.000 (100,000)
1 50,000 52,300 0.877 45,867
2 50,000 54,706 0.769 42,069
3 50,000 57,222 0.675 38,625
Net present value 26,561*

Real method

Table 9.12

Year Real cash flow $ Real discount factor @9% Present value $
0 (100,000) 1.000 (100,000)
1 50,000 0.917 45,850
2 50,000 0.842 42,100
3 50,000 0.772 38,600
Net present value 26,550*

Real discount rate = 9 per cent (1.14 ÷ 1.046 = 1.09)


The small difference of $11 between these two calculations (Tables 9.11 and 9.12) is due simply to rounding.

Exercises: now attempt Exercise 9.2 on page 364

Annuities
If the level of cash flow from an investment is the same from year to year, it is referred to as an
annuity. If this is the case, there is an easier way of calculating the NPV by using annuity tables.
Annuity tables show the present value of $1 received every year, starting one year from now and
going on for n years. A set of annuity tables is available in Appendix E.

WORKED EXAMPLE 9.9 Annuity


A company is considering investing in a project which will cost $10,000 and will yield cash inflows of $3,000 per year
for five years. No scrap value is expected at the end of the project and the company uses a discount rate of 12 per
cent to evaluate investments. Should the investment be accepted?

Answer:

Table 9.13

Year Cash flow $ Discount factor Present value $


t0 (10,000) 1.000 (10,000)
t1–t5 3000 3.605 10,815
NPV 815

As the investment gives a positive NPV of $815, it should be accepted.

Capital rationing: the profitability index


Capital rationing refers to a situation in which there is a limited supply of capital to finance
investment projects. In this situation, a company cannot accept all projects with positive NPVs if
the costs of implementation will exceed the supply of capital. The company will therefore need
to choose between alternative investments.
In practice, capital rationing arises for one of two reasons:

Hard rationing: capital markets will always supply a limited amount of capital. A company
therefore may not be able to raise sufficient capital to finance all available projects.
Soft rationing: the company may have sufficient funds available but has chosen to restrict its
capital investment for strategic reasons.
When faced with a situation of capital rationing a company should allocate available capital so as
to maximize returns on the capital invested. Individual investment proposals should therefore be
considered in terms of their rate of return (ie NPV divided by capital required). This ratio is
sometimes known as the profitability index:

WORKED EXAMPLE 9.10 Capital rationing


A company has $10 million available to fund new projects in the current year. It has identified five potential
investments and calculated the NPV on each investment as shown in Table 9.14.

Table 9.14

Investment NPV Capital required


A $2m $4m
B $1m $3m
C $0.4m $0.5m
D $0.5m $0.75m
E $1.6m $4m
Total $12.25m

The company cannot undertake all five projects because this would require a total capital commitment of $12.25
million. In order to choose where to invest the $10 million available, the company will rank the proposals according to
their rate of return, ie the ratio of NPV/Capital required (Table 9.15).

Table 9.15

Investment NPV/Capital required Ranking


A $2m/$4m = 50% 3rd
B $1m/$3m = 33% 5th
C $0.4m/$0.5m = 80% 1st
D $0.5m/$0.75m = 67% 2nd
E $1.6m/$4m = 40% 4th

Based upon this ranking the company would invest in projects C, D, A and E in that order of preference. This would
use a total capital of $9.25 million. Although this leaves $0.75 million unused, it is unlikely that project B could be
subdivided such that 0.75/3.0 or 25 per cent of the project could be undertaken.

Replacement decisions
All of the investment scenarios we have looked at so far have involved investing in new assets.
However, in reality, many investment decisions involve replacing existing assets. An important
decision for a company is how long to retain an asset such as machinery or a company vehicle
before replacing it. It is possible to use NPV calculations to determine the optimum time at
which to replace an asset.
When using NPV for replacement decisions the technique is modified slightly. Firstly, the
NPV is calculated in the normal way, and then the figure is adjusted to determine the ‘annualized
NPV’. The company should then replace an asset at the point when the annualized NPV is
maximized.

WORKED EXAMPLE 9.11 Replacement decisions


A company purchases a machine for $15,000. The machine can be used for up to three years before it will be
replaced by an identical machine which will be used in the same production process. The following figures have been
estimated (Table 9.16).

Table 9.16

Year 1 2 3
Net revenue $ 9,000 7,500 4,500
Scrap value $ 6,000 4,500 1,500

The company uses a discount rate of 10 per cent for project appraisal. Should the machine be replaced after one, two
or three years?

Answer
If the machine is replaced at the end of year 1, the NPV will be as shown in Table 9.17.

Table 9.17

Year t0 t1 NPV
Cost of machine (15,000)
Net revenues 9,000
Scrap value 6,000
Total (15,000) 15,000
Discount factor 1 0.909
Present value (15,000) 13,635 (1,365)

If the machine is replaced at the end of year 2, the NPV will be as shown in Table 9.18.

Table 9.18

Year t0 t1 t2 NPV
Cost of machine (15,000)
Net revenues 9,000 7,500
Scrap value 4,500
Total (15,000) 9,000 12,000
Discount factor 1 0.909 0.826
Present value (15,000) 8,181 9,912 3,093

If the machine is replaced at the end of year 3, the NPV will be as shown in Table 9.19.
Table 9.19

Year t0 t1 t2 t3 NPV
Cost of machine (15,000)
Net revenues 9,000 7,500 4,500
Scrap value 1,500
Total (15,000) 9,000 7,500 6,000
Discount factor 1 0.909 0.826 0.751
Present value (15,000) 8,181 6,195 4,506 3,882

Each of the NPVs is then ‘annualized’ by dividing it by the annuity factor for the number of years the investment runs:

Replacement after two years gives the highest annualized NPV. This is therefore the replacement strategy which the
company should follow.

Investment appraisal within context


So far in this chapter we have looked at the main techniques of investment appraisal and how
they can be applied to different investment decisions. We will now look at some of the practical
considerations of investment appraisal and some of the problems involved in applying these
techniques in practice.
There are a number of problems to be found in the way these investment appraisal techniques
are actually used in practice. We will focus on three main aspects of practice:

Firstly, the techniques we have examined have a very narrow focus on tangible factors which
can be quantified financially. In practice, many investment decisions can involve factors
which are difficult to quantify or which are intangible, for example investing in a new
computer system in order to improve customer service, or investment in research and
development for new product features.
Secondly, the highly computational nature of the techniques, particularly NPV, means that, in
practice, managers can get lost in the details of the calculations and lose sight of the fact
that they are based upon forecasts which may not be accurate and which in all likelihood
will not actually work out.
Finally, the narrow focus of traditional investment appraisal techniques (typified by the
concept of relevant cash flows) means that they can fail to capture the richness of
investment decisions within the context of a wider organizational strategy. In reality,
managers will not just invest for immediate financial return, but also for wider strategic
reasons such as increasing operational flexibility, gaining competitive advantage or
providing more strategic options in the future. Furthermore, within today’s business context,
managers must consider more than just immediate financial returns to the business’s
owners; the interests and requirements of other stakeholders such as employees, customers,
suppliers, wider society and the environment can be equally important in investment
decisions.

Integrating qualitative factors


One of the problems with traditional investment appraisal is that it tends to lead to an analysis of
tangible financial issues in isolation from other important elements of the investment decision.
When qualitative considerations are brought into an appraisal, the exercise becomes more
complex, even if all the elements being considered are still tangible.
Let us take the example of a company that is investing in a new computerized customer
management system (CMS). This new CMS will be integrated into the existing IT-based
financial system and will be available in all offices through a new computer network. When
doing a financial investment appraisal, there are a number of questions which may be difficult to
answer:

What tangible costs should be included? The tangible costs of new networks and other
computer hardware can easily be identified. However, what is not so easy to establish is the
proportion of these costs which should be allocated to the CMS investment. The network
will be of benefit to all departments and all systems, and any new computers or
workstations will be used for a number of other functions.
How should intangible costs be measured? Some of the costs of implementing the new
CMS system will be intangible and difficult to measure. For example, management time
spent on developing and reviewing the new system will be a distraction from other
activities; there may be disruption to existing work as a new system is implemented; staff
learning to use a new system will be slower and less productive. In such cases it may be
possible to record and measure the amount of time spent on the project, but it is not so easy
to record and measure the impact which this has on other activities and productivity levels.
Diminishing returns from benefits. If the rationale for implementing the new CMS is to
improve customer services, then how far should this go? More and more features could be
added to the CMS at an ever-increasing cost. How many of these features should be
adapted? In a similar manner, increased processing power will provide information faster. It
may be possible to identify that there is a clear advantage in obtaining information in two
hours rather than two days. But would it be better to obtain that information in two minutes,
or even in two seconds?
How should intangible benefits be measured? The problem with implementing any new
investment (particularly an IT investment) aimed at improving the availability of
information or quality of service is that the costs are usually tangible and easily identified,
whereas the benefits are intangible. Traditional NPV calculations are not able to capture
such intangibles, which means that any investment appraisal will inevitably result in a
negative NPV. In many cases, the identification of costs is much easier than the
identification of benefits. This is particularly the case where the benefits sought from an
investment are intangible. There are therefore three options available in this case:
1 Ignore the benefits: the traditional approach is to ignore intangibles and focus only on that
which can be measured in financial terms. Unfortunately, this could result in negative NPVs
and the rejection of investments, simply because the benefits are not immediately
financially quantifiable.
2 Quantify the benefits: a common approach is to attempt to quantify the benefits so that
they can be incorporated into a traditional investment appraisal calculation. Although this
may work in some cases, it may be extremely difficult to quantify some benefits or it may
result in a loss of richness of information about the benefits of investment.
3 Change the approach: the balanced scorecard. A third alternative is to move away from a
purely financial analysis to one which incorporates non-financial benefits. This approach
tries to integrate qualitative and quantitative dimensions of evaluation within one exercise
and can be termed a balanced scorecard approach. To use such a balanced scorecard
approach in investment appraisal requires the decision-maker to weigh up the various
quantitative and qualitative aspects of the investment and to give a ‘score’ to each aspect so
that each investment can be given an overall score, which allows comparison between
alternatives.

WORKED EXAMPLE 9.12 Balanced scorecard approach


Investment appraisal of new customer management system:

Financial benefits:

NPV $x.

Non-financial benefits:

x per cent of users happy with quality;


x per cent of users happy with scope;
x hours saved per month.

Exercises: now attempt Exercise 9.3 on page 365

Addressing risk: the variability in outcomes


In investment appraisal, risk is about the variability of outcomes. When we talk about investment
risk we mean that the actual outcome may not be as we expect. In all the examples that we have
looked at so far, calculations have been based upon one ‘best guess’ set of outcomes.
Unfortunately, in reality, ‘best guess’ will hardly ever be the actual outcome. Therefore, one way
of further incorporating risk assessment into investment appraisal is to analyse the range of
possible outcomes. There are three main ways of doing this: sensitivity analysis; scenario
analysis; and probability analysis.

Sensitivity analysis
Sensitivity analysis is a simple but powerful technique which is used extensively in practice. It
examines the impact on the project return of changes in individual variables such as the
investment cost, the level of cash flows and the life of the investment. The technique is applied
by adjusting each major variable in an NPV calculation until the NPV equals zero (ie the point at
which the investment is no longer financially viable). This gives the decision-maker an
indication of how much of a change can be tolerated in each variable. If an NPV of zero is
arrived at with a relatively small change in a variable, the investment is very sensitive to the
value of that variable. Sensitivity analysis makes it easier to identify weaknesses in forecasts. It
is an extremely good way of identifying key factors that will need careful monitoring once the
project has commenced. It could also be used to evaluate whether further action should be taken
to minimize risk, for example interest-rate hedging or the use of currency derivatives to manage
exchange-rate risks.

Scenario analysis
Whereas sensitivity analysis is useful in isolating and identifying changes in individual variables,
it is not realistic in its presumptions. In reality, it is more likely that several or all variables will
change at the same time, depending upon the circumstances. Scenario analysis addresses this by
predicting how multiple variables will change under different conditions, for example under
improved economic conditions, or if an economic downturn occurs. It is normal for this type of
financial modelling to be done using computer spreadsheets.

Probabilities and expected values


Sensitivity analysis and scenario analysis introduce the idea of using a range of different
forecasts rather than a single calculation. However, they can make decision making more
difficult because they give no indication of the likelihood that each of a range of different
outcomes may occur. A second problem is that a range of different scenarios will give a range of
different NPV values, some of which may be negative. Does that mean the investment should not
be made? It is no longer possible to use the simple rule that a positive NPV means a financially
viable investment.

WORKED EXAMPLE 9.13 Probability analysis


A method to overcome these problems is used by some investment analysts. This involves identifying a range of
different outcomes and attaching a probability to each outcome. These probabilities can be used to calculate a
weighted average of the NPVs of each different scenario as shown in Table 9.20.

Table 9.20

Scenario NPV $ Probability NPV × Prob


A (100,000) 0.2 (20,000)
B 200,000 0.3 60,000
C 300,000 0.4 120,000
D 500,000 0.1 50,000
ENPV = (NPV × Prob) = 210,000

This weighted average of $210,000 is known as the expected net present value (ENPV). It offers the advantage of
once more providing a single figure which can be used to evaluate the investment. If the ENPV is positive, the
investment is financially viable. However, care should be taken in using this technique as the ENPV does not reflect
the outcome which will actually occur. For example, the decision-maker in Worked Example 9.13 should always
remember that there is still a 20 per cent chance that the investment will have a negative NPV of –$100,000.

Other methods of assessing risk


There are a number of other methods which are becoming increasingly popular as means of
modelling the uncertainty around investment cash flows as the availability of sophisticated
computerized spreadsheets and increased processing power has made possible the use of more
complex mathematical techniques.
The Monte Carlo simulation method involves the use of a probability distribution and random
numbers to estimate net cash-flow figures. If this is repeated many times, a distribution of
possible NPVs is derived from which it is possible to ascertain the uncertainty surrounding the
project. Similar methods for modelling uncertainty include the use of Markov chain theory and
fuzzy set theory. These methods allow unknown cash flows to be represented by a range of
inexact (‘fuzzy’) numbers for the purposes of modelling NPV outcomes.

Taking a broader strategic view


So far we have looked at investment appraisal as an isolated financial exercise. The techniques
we have examined, although widely used, have been criticized for not taking into account wider
strategic considerations. In this last section we will therefore look at some recent developments
in investment appraisal which have been aimed at integrating a broader strategic focus into the
financial evaluation of individual investment decisions.

Real options
The traditional investment appraisal techniques we have examined in this chapter take a very
narrow financial focus. This is exemplified by the concept of relevant cash flows, whereby any
cost or revenue deemed not to be relevant to an investment is excluded from the calculation.
However, when looking at an investment from a wider strategic perspective, a business may wish
to build in flexibility which will give it options in the future. For example, if investing in a new
production facility, a company may build a new factory which is twice the size of that needed for
current production capacity. It will do this in anticipation of future growth. The problem with
applying traditional approaches to investment appraisal in this case is that the extra cost of the
large factory in relation to the revenues based upon current capacity may result in a negative
NPV.
The concept of real options borrows from financial options, which are options to exchange a
financial asset such as a share for cash. A real option is the option to exchange a real asset for
some other asset. For example, buying a prime plot of land can give a food retailer a real
expansion option to acquire the revenues from a new outlet by paying the cost of building a new
store on that land. The further option to sell the land in the future, should the new store prove to
be less profitable than anticipated, is a real abandonment option.

Value chain analysis


One way of evaluating a project’s strategic issues as well as its cash flows is to undertake value
chain analysis. This involves identifying strategically important value-creating activities. The
‘value chain’ is that set of activities which link from basic raw materials through to the ultimate
end product. Focusing on these activities involves finding opportunities to enhance customer
value or lower production costs. It has been found in practice that value chain analysis can
produce very different investment decisions from those using traditional techniques, as the
linkages between different activities within the value chain become an important aspect of the
decision process.

Cost-driver analysis
Cost-driver analysis borrows from the concepts of activity-based costing which were examined
in Chapter 8. It involves identifying those cost drivers which flow from the organization’s
investment decisions. By making these connections more explicit, the organization is able to
identify the impact on future cash flows which an investment will make.

Competitive advantage analysis


Competitive advantage analysis involves evaluating whether an investment’s benefits are
consistent with the organization’s competitive positioning strategy, such as cost minimization or
differentiation. Projects can be ranked according to their ability to contribute towards the
organization’s chosen strategy.
Extract 9.4: Cisco – integrating the wider strategic picture
Cisco Systems is a US multinational corporation based in San Jose, California, that designs, manufactures,
and sells networking equipment. During the economic boom of the late 1990s and early 2000s, middle
managers were given great freedom to acquire business start-ups for the technology and ideas. However,
during the economic downturn that followed, Cisco tightened up their investment procedures by creating an
investment review board that met monthly to vet potential acquisitions. Managers proposing acquisitions were
required not only to demonstrate the potential financial benefits but also to draw up detailed integration plans.

Conclusion
This chapter has examined a multitude of investment appraisal techniques. It has presented those
techniques traditionally used for investment appraisal together with some more recent
innovations. In particular, we have evaluated each technique, pointing out its strengths and
weaknesses. The reader will have noted that all the techniques presented in this chapter have
both strengths and drawbacks. It might therefore be concluded that reliance upon one technique
alone may lead to sub-optimal decision making or even to failure. On a practical level, therefore,
it makes sense for an organization to use a mixture of techniques in order to eliminate or
minimize the drawbacks of each individual technique used.
We hope that as a result of studying this chapter the reader will have a better understanding of
how different investment appraisal techniques should be applied, will be able to use them more
effectively to evaluate investments and will be able to identify and avoid common mistakes in
the application of the techniques.

COMPREHENSION QUESTIONS
1. Why is good investment appraisal important to organizations?
2. What are the three main factors that should be considered when appraising a potential investment?
3. What are the main drawbacks of the payback method of investment appraisal?
4. Explain the concept of the time value of money.
5. What is the ‘discount rate’ used in NPV calculations and how is it arrived at?
6. What is an opportunity cost and why should it be included in an investment appraisal calculation?
7. In what ways can taxation impact upon cash flows from an investment?
8. Explain two ways in which risk assessment can be incorporated into investment appraisal.

Answers on pages 366–367

Exercises
Answers on pages 368–372
Exercise 9.1: Basic computations
A company is considering investing in a new production facility at a cost of $120 million. The new facility is
expected to produce annual cost savings as set out in Table 9.21. The facility is expected to have a useful life
of eight years, before becoming obsolete and requiring replacement. The company has a policy of depreciating
all assets on a straight-line basis.

Table 9.21

Year Annual cost savings


1 $30m
2 $35m
3 $40m
4 $45m
5 $30m
6 $26m
7 $15m
8 $15m

Required:
Evaluate the investment using the following techniques:

(a) Payback period – the company considers a capital investment to be acceptable if it pays back within four
years. Should the company make the investment?
(b) ARR – what is the accounting rate of return on the average capital employed?
(c) NPV – if the capital investment has a required rate of return of 12 per cent, what is its net present value?
Should the company make the investment?

Exercise 9.2: Understanding principles


Freshfare Co is a food retailer with 25 stores in the south of the country in which it operates. The board of
directors are currently considering expansion into the north of the country by opening a large new store in a
major northern city.
The investment in the new store is estimated to cost $40m. This will be financed mainly through a new bank
loan of $35m at a cost of 8 per cent a year. The investment is expected to pay back within three years with an
IRR of 22 per cent.
You have been asked to make a presentation on the proposed investment at the next meeting of the board
of directors. The directors have raised some queries regarding the calculations in the investment appraisal and
would like you to address the following points in your presentation:

1. A feasibility study for the new store has already been completed at a cost of $28,000. This cost has not
been included in the investment appraisal calculations.
2. The interest payments on the bank loan of $35m will be payable quarterly. These payments have not
been included in the investment appraisal calculations.
3. The chief accountant proposes to charge 5 per cent of central office administration costs to the new
store. This charge has not been included in the investment appraisal calculations.
4. The company has a policy of depreciating all new investments on a straight-line basis over four years. No
depreciation charge has been included in the investment appraisal calculations.
5. When the new store opens, it will be managed by one of the company’s most experienced store
managers. This manager earns $40,000 per year and this cost has been included in the investment
appraisal calculations. When the manager moves to the new store, her assistant manager will be
promoted to take over her current job. The assistant manager currently earns $25,000 per year but will
receive a salary increase to $30,000 per year when he is promoted.
Required:
Make notes for your presentation to the board of directors which explain the treatment of each of the five issues
above. You should state whether you agree with the accounting treatment in each case. If you disagree with
the accounting treatment, you should explain why and propose an alternative.

Exercise 9.3: Financial appraisal within context


ANG Co is a manufacturer of high-performance processor chips for smart phones and other mobile devices.
The company, based in Europe, has grown rapidly over the last five years. It has been able to compete with
global competitors through developing a highly skilled and loyal workforce.
The company forecasts continued growth in existing markets and intends to break into the new markets of
China and South East Asia. With this in mind, the directors have been examining options for opening a new
factory within the next 18 months.
The directors have identified three possible new sites for the factory. You have been appointed as a
business consultant to help the business choose which site to develop.
The financial information for each factory is set out in Table 9.22.

Table 9.22

Factory A Factory B Factory C


Initial cost $150m $150m $140m
Expected production life 5 years 5 years 4 years

The company accountant has calculated the information shown in Table 9.23.

Table 9.23

Factory A Factory B Factory C


Payback period 3 years 2 years 2 years
Accounting rate of return 29% 29% 32%
Net present value $25.6m $39.4m $28.7m

The NPV is calculated using the company’s standard discount rate of 13 per cent. The following further details
are provided:

Factory A: This factory will be opened next to the existing factory. This will provide more jobs for people in
the area and possible promotion opportunities for existing employees.
Factory B: Factory B will be located in a new enterprise development zone which is situated approximately
150 kilometres from the existing factory. By opening the factory here, the company can take advantage
of some generous tax breaks and other incentives offered by the government. Opening this factory will
involve moving some of the existing production into the new factory. This will mean making 20 per cent
of the existing workforce redundant. (The cost of redundancies is built into the figures above.)
Factory C: This factory will be opened in China. The company will benefit from cheaper labour costs (this is
built into the figures above). The company will also be in a strong geographic position to grow sales in
the newly opened market in China and South East Asia.

Required:
Write a report to the directors of ANG Co which evaluates each of the three potential investments using the
financial and non-financial information provided above. State what further information the directors might need
to consider before making a final decision.
Answers to comprehension questions
1. Good investment appraisal is important in ensuring that new investments offer acceptable financial
returns and that they contribute towards achieving organizational strategic objectives.
2. The three main factors that should be considered when appraising a potential investment are risk, return
and timing.
3. The main drawbacks of the payback method are that:
– it ignores cash flows outside of the payback period;
– it provides no information about total return;
– it does not differentiate between different timing of cash flows within the payback period.
4. The time value of money refers to the fact that when payments are made or received has an impact upon
their value to the investor.
5. The discount rate used in NPV calculations is the percentage rate at which the estimated cash flows from
the investment are discounted. The discount rate will be arrived at by incorporating the rate of return from
the investment required by investors (this must exceed financing costs) adjusted for inflation and the risk
of the investment. Investment perceived as being more risky will be evaluated using higher discount
rates.
6. An opportunity cost is the cost of an alternative course of action which must be forgone in order to
undertake the investment in question. Opportunity cost is measured in terms of financial contribution. It
should always be included in investment appraisal calculations to reflect the fact that when resources are
limited there is always a cost of diverting them from elsewhere.
7. Taxation impacts upon cash flows from an investment in two main ways: taxation paid on earnings from
an investment is a cash outflow; and capital allowances on assets purchased for an investment will
reduce the amount of taxation paid by the business.
8. Risk assessment can be incorporated into investment appraisal in a number of ways. These include:
– adjusting the discount rate used in NPV calculations;
– performing sensitivity analysis on investment appraisal calculations;
– undertaking scenario analysis to evaluate a range of different possible outcomes;
– assigning probabilities to different possible outcomes and calculating expected values;
– using mathematical models such as the Monte Carlo simulation method or fuzzy sets.

Answers to exercises

Exercise 9.1: Basic computations


Payback period
The payback period can be seen by calculating the cumulative cash flow from the investment (Table 9.24).

Table 9.24

Year Cash flow $ Cumulative cash flow $


0 (120) (120)
1 30 (90)
2 35 (55)
3 40 (15)
4 45 30
5 30 60
6 26 86
7 15 101
8 15 116

The cumulative cash flow shows that the investment pays back within the fourth year. If it is assumed that the
cash flow arises evenly throughout the year, the payback period will be:

3 years + (15/45 × 12 months) = 3 years and 4 months

This payback period is within the company’s requirement of four years. The investment should therefore be
accepted.

Accounting rate of return


Total cash inflow over the eight years of the project = $236m
Total depreciation = $120m (the investment will be fully written off by the end of year 8)

Therefore average profit = ($236m – $120m)/8 years = $14.5m

As the investment will be fully written off by the end of year 8, average investment = ($120m + $0)/2 = $60m.

Therefore, ARR = $14.5m/$60m = 24.2 per cent

Net present value

Table 9.25

Year Cash flow $m Discount factor (12%) Present value $m


0 (120) 1.000 (120.00)
1 30 0.893 26.79
2 35 0.797 27.90
3 40 0.712 28.48
4 45 0.636 28.62
5 30 0.567 17.01
6 26 0.507 13.18
7 15 0.452 6.78
8 15 0.404 6.06
Net present value 34.82

The NPV is $34.82m. As this is a positive figure, the investment should be accepted.

Exercise 9.2: Understanding principles


1. Because the feasibility study has already been completed it is a sunk cost. Only future costs which will
change as a result of the decision to invest should be included in the investment appraisal calculation. It
is therefore correct to exclude this cost from the calculations.
2. The cost of financing an investment should not be included within the cash-flow forecasts. It is therefore
correct to exclude this cost. When using the IRR technique, financing costs are considered by evaluating
whether the predicted IRR is sufficient to meet those costs. In this case the predicted IRR is 22 per cent,
which is substantially in excess of the 8 per cent cost of the loan. In any case, it is incorrect to evaluate
an investment purely against the source of financing. The investment must meet the overall return
requirements of the business and not just the cost of the loan to finance the store.
3. Unless it is expected that central office administration costs will increase as a result of opening a new
store, these costs should not be included, as they will be incurred anyway, even if the investment does
not go ahead. However, any additional central office costs which are incurred as a result of opening the
new store should be included within the investment appraisal calculations.
4. Although it is important to charge depreciation for financial accounting purposes, this charge is not
relevant to the investment appraisal. The two methods used to evaluate the proposed investment are
payback period and IRR. Both of these techniques use cash flows and depreciation is an accounting
adjustment, not a cash flow.
5. This is an example of an opportunity cost. It is incorrect to include the salary of $40,000 of the manager
in the investment appraisal calculations, because this salary will be paid regardless of whether or not the
new store is opened. The only additional cost which the company will incur as a result of opening the new
store is the increase in the salary of the assistant manager. The annual cost included in the investment
appraisal should therefore be $5,000, the value of the assistant manager’s salary increase.

Exercise 9.3: Financial appraisal within context


Consultant’s report to the directors of ANG Co

Investment appraisal for the opening of a new factory


Terms of reference
I have been asked to advise the board of directors in relation to its decision to open a new factory in order to
extend its production facility. This report sets out my evaluation of the options currently being considered by the
board.
Introduction
ANG Co has grown rapidly over the last five years. The company is now seeking to extend its production facility
through opening a new factory. In particular, the company intends to break into new markets in China and
South East Asia. Any new factory must therefore meet the needs of the company in developing this new
market.
Three alternatives have been proposed for the siting of the new factory. The first alternative (Factory A) is to
site the factory alongside the existing production facility. The second alternative (Factory B) is to move to a new
enterprise development zone approximately 150 km from the existing factory. The third alternative (Factory C)
is to open a new factory in China.
Financial evaluation
The cost of each of the investments is very similar. Both of the factories in the existing country will cost $150
million. The proposed factory in China is slightly cheaper at $140 million. All production facilities are expected
to have similar lives: the two factories in the existing country will have a production life of five years, whereas
the factory sited in China would have a production life of four years.
A financial evaluation of the three options has already been conducted. The potential investments have been
evaluated using three commonly used and reliable techniques: payback period; accounting rate of return; and
net present value. However, the NPV calculations have all been performed using the company’s standard
discount rate of 13 per cent. This has therefore not been adjusted to take account of the differing levels of risk
involved with each of the three investments. It is therefore important to consider risk and other non-financial
factors alongside the financial evaluations which have already been conducted.
Factory A
This option appears the least attractive in terms of the financial evaluations. It has the longest payback period
of three years. The ARR is 29 per cent, which matches that of Factory B but is less than the 32 per cent return
offered by Factory C. This investment has the lowest NPV of $25.6 million.
However, there are a number of non-financial factors which, when considered, may make this option more
attractive. The current success of the business has been built upon a highly skilled and loyal workforce. By
building the new factory alongside the existing production facility, the company will be able to take advantage of
the skills of the existing workforce. It would also provide an opportunity for promotion and development of
existing staff. This is likely to maintain or increase staff morale and enable ANG Co to retain skilled and
experienced staff. These factors make Factory A a lower risk option than the other alternatives, and if the
discount rate were adjusted to reflect this then the NPV would appear more attractive in comparison with
Factory B and Factory C.
Factory B
This is currently the most financially attractive option as it offers the highest NPV of $39.4 million. Although the
ARR of 29 per cent is lower than that offered by Factory C at 32 per cent, the NPV is a more reliable measure
of the financial viability of the investment.
However, these financial figures should be tempered with consideration of other factors. This option would
involve moving a substantial part of production from the existing facility and making 20 per cent of the existing
workforce redundant. This is likely to have a significant negative impact on the morale of remaining employees.
This may have an impact upon employee motivation and efficiency and may increase staff turnover. The
company would also be losing potentially valuable skilled employees. As the new factory would be opened a
substantial distance from the existing factory, new employees would have to be recruited. It will be more
difficult to supervise and train these new employees alongside existing staff. This will increase the difficulty and
timescale of establishing a skilled workforce of the calibre of the existing factory. The company also needs to
consider the risks of relying upon tax breaks in a new enterprise development zone. There is a risk that the
government may reduce or withdraw these breaks after a short period and this would have a negative impact
upon the projected financial figures.
Factory C
This option is financially attractive. It offers the highest ARR, a two-year payback period and a reasonable NPV
of $28.7 million. It also requires a slightly lower investment than the other two alternatives, which would save
the company $10 million in setup costs. However, the expected production life of this factory is only four years,
against a five-year life from the first two options.
The major advantage of this option is that it would put production on the doorstep of the new market which
ANG Co is seeking to develop. This may offer several advantages in terms of supply time and reaction to
changes in market demands. However, in terms of management control, this option carries the highest risk.
The company currently has no experience of operating a production facility in China. The new factory would be
at a greater geographic distance from the existing business, which will increase problems of control and
communication. The company would also need to recruit a new workforce in an unknown market. This means
that the quality and reliability of employees in the new factory will be unknown.
As this option carries the highest risk, the discount rate used in the NPV calculation should have been
increased. If the NPV were to be recalculated using a higher discount rate, this option might appear
substantially less financially attractive.
Recommendations
Each of the three options offers different strategic advantages to ANG Co. However, they all carry very different
risks. Before a final decision is made, it is recommended that the NPV calculations are adjusted to take account
of these different risks.
10
Operational decisions

OBJECTIVE
To enable the manager to incorporate financial evaluation into their operational decision making and problem solving.

LEARNING OUTCOMES
After studying this chapter, the reader will be able to:

Assess the financial consequences of a range of decision-making situations.


Define the scope and limitations of the financial techniques applied.

KEY TOPICS COVERED


The chapter will examine a range of operational decisions and the financial techniques which can be used to support them:

setting sales targets;


predicting the impact of price changes;
outsourcing vs in-house operation/production;
operational restructuring/automation of business processes;
closing a business segment;
dropping a product/service line.

MANAGEMENT ISSUES
Managers need to be able to assess the financial impact of operational and tactical decisions.

Introduction
In this chapter we will look at some financial aspects of operational decision making. We
introduce some fundamental principles of financial decision making and some core decision-
making techniques. Once these techniques are understood, we examine their wider applications
and demonstrate their use in a range of typical business decision-making situations. A key skill
for managers is understanding which is the most appropriate technique for a given decision.

Operational decision making


All managers are faced with operational decisions. These are decisions concerning the best
means of implementing strategies and overcoming problems that are encountered. There is a
saying that ‘analysis is at the heart of business intelligence’: in order to be successful in dealing
with the issues they face, managers need a systematic approach towards decision making
together with appropriate tools for evaluating the financial impact of different options.
A typical operational decision making process will involve a number of key steps as
illustrated in Figure 10.1.

Figure 10.1 Decision-making steps

These steps involve:

Step 1: recognizing and defining the problem or issue which needs to be addressed.
Step 2: identifying alternative solutions to the problem and eliminating alternatives that are
not practically feasible.
Step 3: identifying the costs and benefits of alternatives in order to establish their financial
feasibility.
Step 4: assessing qualitative factors in the decision that go beyond the immediate financial
benefits.
Step 5: selecting the alternative which offers the greatest overall benefit when considering
both financial and qualitative factors.

In this chapter we are primarily concerned with step 3: identifying the costs and benefits of
alternatives. However, we will also visit some of the qualitative factors in step 4 that will
impinge upon decisions.
We will look at a number of typical operational decisions. In order to examine these different
business situations we will explore two key areas of financial decision-making theory in this
chapter: cost–volume–profit analysis (CVP) and relevant costing. We will introduce each theory
in turn and then demonstrate its application through a number of examples which relate to real-
world business decisions.

Cost–volume–profit analysis (CVP)


CVP analysis is the study of the interrelationship between costs and revenues (and therefore
profit) at various levels of activity.
Economic models which measure changes in costs and revenues as the volume of activity
increases can be complex. However, for the purpose of managerial decision making it is possible
to simplify these models in a way that makes them easy to use and therefore more readily useful
to the average manager. CVP analysis therefore makes a number of assumptions about how
revenues and costs behave as the volume of activity within a business increases. We will look at
each of these assumptions in turn.

Revenues
Economic models tell us that as the volume of sales increases, the unit sales price will decrease.
Although this may be true when looking at the full spectrum of possible levels of sales, when
using CVP analysis for decision making we are usually looking at a relatively narrow range of
sales volumes. It is therefore possible to make the assumption that unit sales price will remain
constant across all levels of activity. In this case we can say that total sales revenue will be the
volume of sales (in units) multiplied by the unit sales price:

Sales revenue = Units sold × Sales price

This relationship between sales revenue and the volume of activity (units sold) is represented
graphically in Figure 10.2.

Figure 10.2 Revenue


Costs
In financial decision making we usually divide costs into two broad categories: fixed costs and
variable costs. This categorization refers to how costs behave in relation to changes in volume of
production and sales.

Fixed costs
Fixed costs are unaffected by changes in the level of activity and therefore remain constant or
‘fixed’ as the volume of activity increases or decreases. An example of a fixed cost is the rent
paid on a shop. The rent must be paid even if the shop sells nothing. If the shop does well and
sales reach a very high level, the same amount of rent will be paid.
This relationship between fixed costs and the volume of activity is represented in Figure 10.3.

Figure 10.3 Fixed costs

Expert view 10.1: Fixed costs


Some ‘fixed’ costs may increase if there is a substantial change in the level of activity. For example, a business
may employ a manager in its production department. The manager will be paid a fixed salary and this will not
change with the level of production. However, if production is increased substantially, for example by moving to
24-hour operations and running two shifts of workers, the business may employ a second manager so that
there is one manager for each shift. In this case the cost is said to be a stepped cost, because it remains
constant over a range of activity (the cost of one manager for one shift), but then increases by a ‘step’ once the
activity goes beyond a certain level (the cost of two managers for two shifts).
A second point to note about fixed costs is that they are deemed to be fixed in relation to the level of activity,
but not necessarily over time. Fixed costs may change over time. For example, the rent paid on a shop may be
subject to an annual review, with an increase each year. Similarly, a manager paid a fixed salary may have an
annual pay rise.

Variable costs
Variable costs are those costs that are directly related to the level of activity and will therefore
change as the level of activity changes. For example, the cost of direct materials will increase
with the number of units produced.
The relationship between variable cost per unit and the volume of activity will vary in
different circumstances. For example, in many cases the variable cost per unit will decrease as
volume increases, owing to economies of scale. This will be the case when direct labour
becomes more efficient as volume increases. It will also be the case if bulk-purchasing discounts
are available when buying large quantities of materials. However, in some cases the variable
costs per unit may increase as volume increases. For example, higher levels of activity may
involve paying employees an overtime premium.
Despite these complicating factors, for financial decision-making purposes we can usually
assume that this relationship is linear, as we did with revenues. That is to say, the variable cost
per unit will remain the same across all levels of activity. This means that variable costs can be
represented graphically, as shown in Figure 10.4.

Figure 10.4 Variable costs

Total cost
The total cost of an activity will be made up of variable costs plus fixed costs. Therefore we can
create a graph to plot total costs by adding these two individual cost elements together. This will
give us a graph, as illustrated in Figure 10.5.

Figure 10.5 Total costs


Now that we have plotted both revenues and costs on graphs in Figures 10.1 to 10.5, we can
incorporate them into one graph as illustrated in Figure 10.6.

Figure 10.6 Total revenue and total costs

The graph in Figure 10.6 shows that up to a certain volume of activity, costs exceed revenue.
However, there comes a volume of activity, known as the break-even point (BEP), at which
total cost equals total revenue. This is the point at which the business makes neither a loss nor a
profit. If the business operates at a volume of activity greater than the break-even point, it will
make a profit. On the other hand, if the business operates at a volume of activity less than the
break-even point, it will make a loss.

Mathematical approach to CVP analysis


Although the relationship between costs and revenues is easy to see and understand using graphs,
it is actually easier to apply to managerial decision making using the following profit equation:
We can rearrange this formula to express profit in a way which is more useful for decision
making:

Profit = Q(P – V) – F

Where: P – V = Contribution per unit

We can use this formula to calculate the break-even point. The break-even point will occur at the
volume of sales (ie the value of Q) at which profit is nil.
The graph in Figure 10.6 illustrates that this lies at the value of Q where total revenue = total
costs. That is to say, where:

(Remember: contribution per unit = sales price – variable cost per unit)

WORKED EXAMPLE 10.1 Computing the break-even point for sales


Isabel wishes to start a business selling ice cream from a mobile van. She can lease a van at a cost of $206 per
week. Isabel has estimated that the van will cost around $70 per week to run. She has also obtained the following
prices from an ice cream wholesaler:

Tub of ice cream – 100 portions $20.00


Box of 100 cones $10.00

If Isabel sells ice creams at $1.50 each, how many ice creams will she need to sell each week in order to break even?

Answer
We can calculate the number of ice creams that Isabel needs to sell by using the break-even formula. First we need
to calculate contribution per unit:
Therefore Isabel will need to sell 230 ice creams per week in order to cover fixed costs and break even.

Target profit
Obviously Isabel will wish to achieve more than simply break-even with her ice cream business.
She will also need to make a profit.
Another application of the CVP technique is to calculate how many units must be sold in
order to achieve a target profit. Let’s do this for Isabel’s business.

WORKED EXAMPLE 10.2 Target profit


Isabel has decided that she needs to earn a minimum profit of $350 per week from her business. How many ice
creams will she need to sell each week in order to earn this profit?

Answer
The break-even formula we used in Worked Example 10.1 told us how many ice creams Isabel needs to sell in order
to cover her fixed costs. If she wishes to earn a profit of $350, Isabel needs to sell sufficient ice creams to cover fixed
costs plus the profit. Hence we can modify the formula we used in Worked Example 10.1 as follows:

Therefore Isabel will need to sell 522 ice creams in a week in order to earn a profit of $350.

Margin of safety
In Chapter 9 we introduced the concept of sensitivity analysis. Sensitivity analysis involves
exploring how susceptible predicted profit levels are to changes in levels of activity. CVP
analysis can be a very useful tool for performing such analysis, and one of the most basic
methods of doing this is to calculate the margin of safety.
The margin of safety is the gap in units of sales between the expected or target sales and the
break-even point. It is a useful measure in that it tells the business how far sales can fall from
their expected level before the business starts to make a loss. This is illustrated in Figure 10.7.
Figure 10.7 Margin of safety

We can apply this analysis of margin of safety to Isabel’s ice cream business.

WORKED EXAMPLE 10.3 Margin of safety


Isabel expects to sell 550 ice creams each week. The break-even level of sales is 230 ice creams per week. What is
Isabel’s margin of safety on her expected sales?

Answer
The margin of safety is 320 ice creams (550 – 230). It is more useful to express this as a percentage of the expected
sales:

MOS = 320/550 = 58 per cent

This means that sales must fall by more than 320 units or 58 per cent from the expected level of 550 units per week
before Isabel will start to make a loss.

The margin of safety is a useful measure of risk. The smaller the margin of safety, the higher the
risk that the business will make a loss if it has a poor week of sales.

Expert view 10.2: Margin of safety and sensitivity analysis


In practice, businesses will usually perform CVP analysis using a computer spreadsheet. The use of a
computer allows for easy manipulation of the figures so that managers can quickly identify how sensitive
predicted profit levels are to changes in different variables.

Change in selling price


Another application of the CVP technique is calculating the impact on profit from changes in
price. In Chapter 8 we looked at the relationship between price and demand. Generally speaking,
if a business reduces its prices then this will increase demand. However, increased demand at a
lower price will not necessarily increase profits. Therefore, before changing prices it is useful to
calculate the predicted impact this will have upon overall profit. We can apply this analysis to
Isabel’s ice cream business.

WORKED EXAMPLE 10.4 Change in selling price


After selling ice creams for several weeks, Isabel has found that she sells 520 ice creams per week on average at a
price of $1.50. Isabel’s fixed costs are $276 per week and her variable costs are $0.30 per ice cream sold.
Isabel is considering reducing the price to $1 per ice cream in order to increase sales.

(a) What is the minimum number of ice creams that Isabel must sell each week in order to justify the reduction in
sales price?
(b) What will be the impact on Isabel’s profit if sales increase to 750 units per week at the new price of $1?

Answer
Isabel currently sells 520 ice creams per week at a price of $1.50. Her weekly profit is therefore:

In order to justify a reduction in the sales price Isabel must make at least the same level of profit at the new price of
$1.
At the new price the new contribution per unit will be: $1.00 – $0.30 = $0.70. Therefore the minimum level of sales
(in units) will be:

Isabel must sell at least 892 ice creams each week at the price of $1 in order to make the same level of profit that she
currently makes selling the ice creams at $1.50. This represents increasing sales of 372 units (892 – 520) or 71.5 per
cent. This analysis shows us that Isabel would need a substantial increase in sales at the lower price in order simply
to make the same level of profit. Unless she can be confident of achieving such a large increase in sales, Isabel
should not reduce the price to $1. Otherwise she runs the risk of reducing her profits:

If Isabel’s sales increase to 750 ice creams per week at the new price of $1, her profit will fall to $249 per week:

Change in production systems


As well as analysing the impact of price changes on profitability, the CVP technique can be used
to analyse the impact of changes in production costs. This is particularly useful for businesses
that are making operational changes. For example, the automation of activities is likely to
increase fixed costs and reduce variable costs. Because this changes the operating gearing
(the mix of variable and fixed costs), it will have an impact upon the break-even point, the
margin of safety, and profitability at different levels of activity.

Expert view 10.3: Operating gearing


Operating gearing (sometimes called ‘operating leverage’) describes the relationship between an organization’s
fixed costs and variable costs:

A greater proportion of fixed costs means a higher operating gearing. Increasing operating gearing makes a
business’s profits more sensitive to a change in volume of activity. If volume increases, profitability can be
greater if operating gearing is higher. However, the effect reverses when volumes fall (see Worked Example
10.5). It is for this reason that businesses with high operating gearing often have to take drastic cost-cutting
measures when sales start falling. Managers must be aware of the impact of operating gearing on profits and
risk to make good business decisions.

WORKED EXAMPLE 10.5 Change in production systems


Angus Co makes beds. Current production information is:

Materials costs $80 per unit


Labour costs $160 per unit
Variable sales costs $40 per unit
Sales price $380 per unit
Fixed costs $1,100,000 per year
Current volume of output 14,000 units per year

The company is considering automating the metal-cutting process, which is the most labour-intensive part of
production. This will involve installing a new computerized cutting machine that will have an annual fixed cost of
$500,000. However, it is expected that with the new automated cutting, labour costs will fall to $120 per unit.

(a) Compare the annual profit using the current production method and using the new automated production, if
annual sales remain at 14,000 units.
(b) Compare the annual profit using the current production method and using the new automated production, if
annual sales were to fall to 12,000 units.

Answer
If sales remain at 14,000 units per year:

At the reduced level of sales (which represents a fall of less than 15 per cent in sales), the proposed new production
method would result in the company earning $20,000 less profit. Therefore, before it goes ahead with the automation,
the company should examine its sales forecasts in detail in order to be confident that future sales will not fall. If it
cannot be confident of this, it should not change its current production methods.

Advertising campaign
CVP analysis can be used to help managers make decisions regarding discretionary expenditure
such as advertising. For example, a company may plan an advertising campaign in order to
increase its volume of sales. CVP analysis will help managers assess the potential impact on
overall profitability.

WORKED EXAMPLE 10.6 Advertising campaign


Digi Co manufactures laptop computers. The computers sell at $500 and the variable costs of production, sales and
distribution are $350 per computer. The company has fixed costs of $9,200,000 per year. Forecast sales for the
coming year are 70,000 units.
At a recent board meeting the marketing director argued that Digi Co computers offer good value in comparison to
those of competitors. She has therefore proposed an advertising campaign to increase customer awareness of this.
The advertising campaign would cost $350,000 and the marketing director is confident that it would increase sales
volume by 4 per cent.
Should Digi Co undertake the advertising campaign?

Answer
If the company undertakes the advertising campaign, fixed costs will increase by $350,000. Total contribution will also
increase because of the increase in sales volume:

Increase in contribution = (70,000 × 4%) × ($500 – $350) = $420,000

Therefore Digi Co should go ahead with the advertising campaign because the net impact will be an increase in profit
of $70,000 ($420,000 – $350,000).

Evaluation of the CVP technique


As can be seen from the worked examples above, CVP is an extremely straightforward analytical
tool. This simplicity is both its strength and its weakness. The strength of the technique lies in
the way that it removes many of the complications of the real world in order to provide a sharp
focus on the financial impact of decisions. This makes the technique both simple to apply and
easy to understand. However, the simplicity of CVP analysis has led to criticism. This criticism
focuses mainly on the underlying assumptions. This can be summarized as follows:

CVP analysis departs from accepted curvilinear models of supply and demand used in
economic pricing theory because it uses simple linear formulae for revenues and costs. In
other words, CVP analysis ignores price elasticity of demand and economies of scale.
Supporters of CVP have responded to this criticism by suggesting that although this is true,
the focus of analysis on limited range of activity volumes means that a linear function is a
reasonably accurate proxy of the true economic model.
CVP analysis focuses too much on the short term. It is typically restricted to one accounting
period. It therefore ignores longer-term strategic issues.
CVP analysis usually assumes a single product. The analysis can be used for multiple
products but it is necessary to assume a constant sales mix, ie the proportion of different
products sold remains the same as the overall volume changes. In practice this rarely occurs.
In the real world, changes in volume of sales can result from market conditions that will
affect different products or services to differing degrees. This limits the usefulness of CVP
analysis for businesses offering a range of products or services.
CVP analysis assumes a simple, single-stage manufacturing process in which fixed and
variable costs can be clearly identified. In the real world, manufacturing processes can be
complex, with the interplay of costs being equally so.
CVP analysis assumes that costs can be categorized into either variable or fixed. It does not
allow for costs with more complex behaviour. In the real world, organizations and their
activities are complex. It is often difficult to separate costs into clear classifications of fixed
and variable.
CVP analysis assumes that the forces influencing a business are static rather than dynamic.
In the real world, changes in volume of sales affect pricing, which in turn can change price
elasticity of demand. Changes in volume also affect the cost of materials used in production
and the cost of direct labour. CVP analysis ignores these changes and therefore does not
reflect the reality that makes up a dynamic complicated market.

All of these criticisms are directed at the simplicity of CVP analysis when used in complex
organizations and dynamic business environments. The static nature of CVP analysis means that
it must be used with care within a dynamic and complex market. However, this does not mean
that the technique is totally without validity. CVP analysis takes a snapshot. This is a useful
snapshot provided managers do not think purely in terms of single points in time with nothing
changing. There is often great merit in a tool that cuts through the complexities of the real world
in order to give simple and straightforward measures for decision-makers.

Cost–volume–profit analysis: summary


This section has demonstrated how CVP analysis can be a useful analytical tool for managers.
However, managers must be aware of the limitations of the technique and therefore decide when
it is an appropriate tool.

Relevant costing
The concept of relevant costs and opportunity costs were introduced in Chapter 9 in the context
of investment decision making. In this chapter we will examine these concepts in more detail and
demonstrate how they can be applied to other financial decisions such as: the decision to buy in a
product or service or make it in-house, or the decision to drop a product or close a division or
segment of the business.
As was the case when we looked at investment appraisal in Chapter 9, it is always assumed
when evaluating alternative courses of action that the objective is to maximize the present value
of future net cash inflows.

Measuring relevant costs


The relevant costs for decision making are those future costs that will be affected by the decision.
Costs that are independent of the decision are not relevant and should not be considered when
making that decision.
As a general rule, a cost that is avoidable by choosing one alternative over another is a
relevant cost. A cost that is unavoidable, no matter which alternative is chosen, is irrelevant and
should not be included in the evaluation of the decision. Figure 10.8 sets out a decision tree for
deciding whether a cost is relevant to a particular decision.

Figure 10.8 The relevant cost decision

The first question to ask is: ‘Has the cost already been incurred?’ If a cost has already been
incurred, it cannot be avoided no matter what decision is made. Such a cost is known as a
sunk cost; it is irrelevant to any decision and should be ignored. This sounds simple in
principle, but often people are confused about the relevance of sunk costs and include them
in their considerations.
For example, a manager may argue that because his organization has invested $500,000
in a new computer system, this expenditure would be wasted if the computer system were
abandoned because of problems. Such a manager would argue that the $500,000 is a
relevant cost of abandoning the system. However, the reality is that the $500,000 has
already been spent, and this cannot change no matter what decision is made now about the
future use of the computer system. The only relevant costs in this case would be those
future costs incurred to rectify the problems with the system.
Secondly, you should ask: ‘Are you committed to incurring the cost?’ In some cases, costs
may not yet have been incurred, but the business is committed to those costs no matter what
decision is made. If this is the case, the cost is irrelevant and should not be included in the
decision-making process.
For example, a business may enter into a five-year lease of a building. After one year the
business thinks that it no longer needs that building and faces the decision of whether or not
to move out. If the business is contractually obliged to continue paying the rent for the
remainder of the five-year lease term, any future rent will have to be paid no matter what
decision is made about the use of the building. The rent cost therefore becomes irrelevant to
the decision, even though it has not yet actually been incurred.
Finally you should ask: ‘Is the cost a real cash cost?’ Some costs are valid accounting costs
that should be included on the profitability analysis, but they are not relevant for decision
making. This is the category of costs which relate to accounting adjustments rather than
direct cash payments. The main example is depreciation. Each accounting period, a cost will
be recorded for the depreciation of non-current assets. However, there is no underlying cash
payment relating to this depreciation cost. It is therefore not a relevant cost.
For example, a business owns a machine which is depreciated on a straight-line basis at
$5,000 per year. If the business decides to sell the machine, it will no longer incur the
depreciation cost of $5,000. However, this is not a relevant cost and should not be included
in the decision regarding selling the machine. The reality is that the underlying cash flow
represented in the depreciation charge was the initial purchase of the machine. As this
occurred in the past, it is a sunk cost and is irrelevant to the decision. The only costs
relevant to the decision to sell the machine would be any extra costs incurred in selling it
(such as advertising) together with the revenue received from the sale.

Using relevant costing for decision making


There are two basic steps in using the relevant costing approach to evaluate an operational
decision: firstly, disregard costs and revenues that are not relevant to the decision; secondly, use
the costs and revenues that remain (ie the relevant costs) to evaluate alternatives, choosing that
alternative which offers the greatest net benefit.
Costs do not rigidly fit into the category of ‘relevant’ or ‘irrelevant’. What is a relevant cost
will depend upon the circumstances of the decision: costs may be relevant in the context of one
decision but not in a different situation. For example, the timescale of the decision often impacts
on which costs are relevant. We will look at this issue in more detail below when we compare
short-term and long-term decisions for outsourcing.
Let us look at how this principle is applied within the context of common operational
decisions.

The make or buy decision


Organizations are regularly faced with what is commonly known as the ‘make or buy’ decision.
This is the decision of whether to make a product (or provide a service) in-house, or to buy that
product in from some external supplier. There can be many advantages of buying in from an
external supplier:

Management is freed up to concentrate on activities which may be more important to the


profitability of the organization.
Risks are transferred outside of the organization (to the supplier).
External suppliers can specialize and create economies of scale which enables them to
supply at a much lower cost than the organization is able to achieve.
Buying in from multiple suppliers creates greater security of supply.
Buying in services when needed may give an organization greater flexibility and save costs
during times when those services are not needed.

During the 1980s the concept of outsourcing became extremely popular. As a result, many
businesses started to buy in from external suppliers those facilities which had previously always
been provided internally.

Extract 10.1: Tom Peters – ‘stick to the knitting’


One of the best-known proponents of outsourcing is the management guru Tom Peters, who coined the phrase
‘stick to the knitting’. Peters’ 1982 book In Search of Excellence (written with Robert Waterman) is one of the
biggest-selling business books ever. Based upon a study of 43 of the world’s most successful (at the time)
businesses, the book identifies eight common themes to that success. One of those themes was ‘stick to the
knitting – stay with the business you know’. One interpretation of this principle, which Peters developed in his
later work, was the idea that management should concentrate on their core activities and let others take care of
the non-core activities by outsourcing them.
Outsourcing became popular and was widespread practice throughout the late 1980s and into the 1990s. It
was seen as a way of keeping businesses lean and flexible – non-core activities could be bought in as needed
and at a lower cost than doing the work in-house. Many businesses outsourced payroll services, legal services,
recruitment and advertising. Businesses also outsourced areas of manufacturing to overseas suppliers, finding
that they could take advantage of cheaper (and un-unionized) labour and avoid regulations, high taxes and
other operating costs. Outsourcing also became popular in the public sector, with many areas of public service
such as public transport, health care and local services being bought in from private-sector suppliers.
This is a trend which continues in the 21st century in both the public and private sectors.

Short-term vs long-term situations


When faced with a decision which covers only the short term, it is often the case that many costs
are unavoidable (and therefore relevant to the decision) because reducing or removing those
costs requires measures which take a longer timescale to implement. In the long term, more costs
become avoidable. This means that the outcome of a ‘make or buy’ decision will differ according
to whether a short-term or long-term position is considered. This principle is illustrated in
Worked Examples 10.7a and 10.7b.
WORKED EXAMPLE 10.7A Make or buy decision – short term
Scott Co operates a distillery producing Scotch whisky. The main ingredient of whisky is malted barley, which is
produced by processing raw barley in order to convert the starch content into sugars.
Scott Co currently produces its own malted barley. However, the directors are considering the option of buying the
malted barley in from an external supplier that specializes in malting.
The estimated cost per tonne to Scott Co of producing malted barley in-house is as follows:

$
Direct labour 80
Direct materials (raw barley) 260
Direct (variable) overheads 180
Fixed overheads (apportioned) 120
640

The outside supplier has quoted a figure of $580 per tonne for an order of 500 tonnes of malted barley. This is equal
to three months’ supply.
The share of fixed overheads apportioned to malted barley production will still be incurred whether or not the
company purchases the malted barley from an external supplier. Any redundancies of production staff will be subject
to three months’ notice.
Should Scott Co continue to produce malted barley in-house or should it buy the malted barley from the external
supplier?

Answer
On face value, the buy-in option appears more financially attractive: it currently costs Scott Co $640 per tonne to
produce in-house, and the external supplier is offering malted barley at $580 per tonne.
However, in evaluating this decision, the directors need to compare the quoted external price of $580 with the
relevant cost of producing the malted barley in-house. The relevant cost will be those costs which can be reduced or
removed if the malted barley is bought in. The fact that this is a short-term contract has an impact on which costs
become relevant.
Because the contract is for three months’ supply and direct labour is subject to three months’ notice on
redundancy, direct labour costs cannot be reduced if the malted barley is bought in. This is therefore not a relevant
cost. Likewise, the fixed overhead costs will still be incurred if the malted barley is bought in and so this is not a
relevant cost. This means that the relevant costs, ie the costs that will be eliminated if the malted barley is bought in,
are:

$
Direct materials (raw barley) 260
Direct (variable) overheads 180
440

These figures reveal that the short-term contract is not financially viable for Scott Co. The company would be
incurring purchasing costs of $580 per tonne, but only reducing its in-house costs by $440 per tonne. This means that
the buy-in option would cost the company $140 per tonne more than manufacturing themselves.
However, the situation may change if the company is able to negotiate a long-term contract for supply of malted
barley.

WORKED EXAMPLE 10.7B Make or buy decision – long term


Scott Co is able to negotiate a long-term contract with the outside supplier to deliver 2,000 tonnes of malted barley
per year for five years at a cost of $580 per tonne. In the long term the fixed overheads and labour costs can be
reduced (with no redundancy costs).
Should Scott Co enter into the longer-term contract for external supply of malted barley, or should it continue to
make the malted barley in-house?

Answer
Because Scott Co is able to eliminate both the fixed overhead costs and the labour costs with a long-term contract, all
the costs of production become relevant. This means that Scott Co would be saving $640 per tonne by not
manufacturing against the cost of buying in of $580 per tonne. This is a net saving to the company of $60 per tonne.
In this case, the option of buying in rather than making in-house appears more financially attractive. The final decision
should, of course, be made after evaluating wider strategic and qualitative factors. (These are discussed below.)

Extract 10.2: Outsourcing at TagHeuer


TagHeuer, the Swiss watch manufacturer, has a clear ‘make or buy’ strategy. The company is the world’s
fourth-biggest producer of luxury watches, after Rolex, Cartier and Omega. Despite this, the company
manufactures very little. Watch movements and other components are bought in from external suppliers. Some
watch cases are made by a subsidiary, but about half are bought in. Even the final assembly of watches is
subcontracted. Only the top-end watches are manufactured internally and that is by a subsidiary.
The company advocates many advantages in this strategy: contracting out manufacturing transfers risk to
suppliers; sourcing from multiple suppliers creates security of supply and competitive pricing; management can
concentrate on product development and marketing.

Opportunity costs
The concept of an opportunity cost was introduced in Chapter 9. This is a category of cost that
will not be found in an income statement. It is used only in decision making. However,
opportunity costs can be extremely important in assessing the true financial impact of a decision.
This is because, if an organization is operating at full capacity, the decision to engage in one
particular activity (for example, to produce a particular good or service) has a cost in terms of not
being able to use that capacity for some other activity.

Expert view 10.4: Opportunity cost


An opportunity cost is a measure of the opportunity that is lost or sacrificed when the choice of one course of
action requires that an alternative course of action be given up. An opportunity cost is always measured
financially in terms of lost contribution (Contribution = Sales revenue – Variable costs).

WORKED EXAMPLE 10.7C Opportunity costs


Let us return to the example of Scott Co considering a short-term three-month contract for buying in malted barley
(see Worked Example 10.7a).
Scott Co is currently running at full capacity. The production of malted barley internally required 20 hours of
production time that could otherwise be used for chill filtering (another part of the whisky production process). Chill
filtering yields a contribution of $8 per hour.
Should Scott Co continue to produce malted barley in-house or should it buy the malted barley from the external
supplier?

Answer
Returning to the analysis that we did in Worked Example 10.7a, we established that the relevant cost of producing the
malted barley in-house was:

$
Direct materials (raw barley) 260
Direct (variable) overheads 180
440

However, because Scott Co is now operating at full capacity, there is also the opportunity cost of using production
time for malting barley rather than chill filtering. This opportunity cost is measured in lost contribution:

20 hours × $8 per hour = $160

When we include this opportunity cost, the total relevant cost of producing malted barley in-house is:

$
Direct materials (raw barley) 260
Direct (variable) overheads 180
Opportunity cost 160
600

This means that the cost of buying in malted barley from the external supplier at $580 per tonne is now more
financially attractive than manufacturing in-house. Scott Co should therefore buy in the malted barley over the three-
month period and use the internal production facility for chill filtering.

The consideration of qualitative factors in outsourcing


The quantitative factors which we have considered above, ie the relevant costs and the
opportunity costs of manufacturing in-house as opposed to buying in, do not give the full picture
in decision-making situations. Other, non-financial factors should always be considered, both
because they are important in their own right, and also because they might have financial effects
which are not immediately quantifiable. Examples of factors which can be difficult to quantify
financially are:

Redundancies. Outsourcing an activity invariably results in redundancies. These can have


an obvious immediate cost. However, there are other costs which can be more difficult to
quantify in terms of lost expertise and knowledge. It is relatively easy to outsource an
activity, but far more difficult to reverse that decision once the internal expertise has been
lost.
Employee morale. Outsourcing which reduces staffing and results in redundancies can have
a negative impact on the morale of remaining employees. This can have a knock-on effect
on productivity levels and on employee retention. High staff turnover carries additional
costs in terms of recruitment and training of new staff and the loss of experience,
knowledge and skills of the employees who leave.
Reliance on suppliers. If goods or services are bought in from an external supplier rather
than produced in-house, the organization becomes reliant upon that external supplier.
Dealing with a third party is always more complex than dealing with issues in-house as it
can involve slower communication, formal contracts and far more bureaucracy.
Production flexibility. External supply usually involves formal contracts which stipulate
quantities of goods or services supplied. Although this can provide the security in times of
stable trading, it can reduce flexibility if the business’s situation changes. For example, if
production and requirements reduce substantially, a business may find itself contractually
bound to purchase large quantities of supplies it no longer needs.
Ability to meet customer requirements. Many industries experience rapid technological
development and change. It is therefore important for businesses to be able to modify and
develop their products to meet the changing demands of customers. If items are produced
in-house, it is more likely that the business will also have in-house research and
development facilities. Therefore, outsourcing, although it may reduce short-term costs,
may also reduce longer-term product development and innovation.
Control over quality. Any organization needs to ensure that it has control over the quality of
what it produces. Often such control is easier if all operations and therefore communications
are in-house. Quality problems can be more difficult to resolve when dealing with external
suppliers because of communication, cultural and contractual issues between the two
separate organizations.

Extract 10.3: Make or buy – did IBM get it wrong?


When IBM launched the PC in 1981 the business was primarily focused on manufacturing computer hardware.
Management were faced with the decision of developing an operating system for the new PC in-house, or
outsourcing. They decided to outsource to Microsoft, a relatively new and small software development
company. If IBM had been able to anticipate the outcome of this decision, they might well have chosen to
develop the software in-house.

The shutdown decision: deleting a business segment


In a changing and developing business environment, organizations are often faced with the
decision of whether or not to delete a segment of the business. This can cover a wide range of
decision-making situations, because a business ‘segment’ could be a:

product;
type of customer;
geographic region;
distribution channel;
or any other identifiable part of a business.

Extract 10.4: Cutting business segments at Nokia


The Finnish company Nokia is known today as one of the world’s leading mobile phone manufacturers.
However, the business started as a paper manufacturer and developed over a 100-year period into an
industrial conglomerate involved in manufacturing paper, tyres, footwear, electrical cables, televisions, personal
computers, consumer electronics, military communication equipment, electronic components, plastics,
aluminium and chemicals.
When the Soviet Union collapsed in the early 1990s, Finland suffered a huge recession and Nokia faced
financial disaster. Management were presented with some challenging decisions and chose to focus on the
rising telecommunications business. All non-telecommunication businesses, which represented nearly 90 per
cent of revenues, were sold off.
In 1991, more than a quarter of Nokia’s revenue came from Finnish domestic sales. However, with the
strategic change and refocusing on mobile phones, Nokia expanded into Europe, North and South America and
Asia. By 1998 Nokia had grown to become the world’s largest mobile phone manufacturer and continued to
hold this position until 2012.

Using relevant costing to evaluate the shutdown decision


The decision to drop a business segment will depend upon the impact the decision has on the
overall profitability of the organization. The business segment should be closed only if this
results in an increase in overall profit. Therefore, in order to assess its impact, it is necessary to
analyse the costs carefully and ensure that only relevant costs are included in the evaluation. This
should be done as follows:

Overall corporate profitability should be analysed by business segment.


In assessing the profitability of individual segments, only those costs that are relevant to the
decision should be included.
The contribution margin that will be lost by closing the segment should be established by
comparing the revenue that will be lost with the costs that will be avoided.

This approach is best seen through the use of an example:

WORKED EXAMPLE 10.8 The shutdown decision


Book Co is a book retailer that has three large stores in three different cities. The budgeted income statement for the
company for the next accounting period is as shown in Table 10.1.

Table 10.1

Store A Store B Store C Total


$000 $000 $000 $000
Sales 1,000 1,800 1,200 4000
Cost of sales 480 720 510 1710
Gross profit 520 1,080 690 2290
Selling costs: 0
Store staff salaries 140 180 150 470
Store running costs 65 95 70 230
Store expenses 25 40 22 87
Advertising 60 60 60 180
Head office expenses 135 243 162 540
Central warehouse costs 120 200 140 460
Total costs 545 818 604 1967
Net profit/(loss) (25) 262 86 323

Books are ordered, stored, and dispatched to individual stores from a central warehouse. It is estimated that 60 per
cent of the costs of the central warehouse are fixed and 40 per cent are variable.
Store expenses are variable, but store staff salaries and store running costs are fixed. However, all store costs are
avoidable if the store is closed. Each store has an advertising budget of $60,000 that is also avoidable if the store is
closed.
The directors of Book Co are concerned that Store A is predicted to make a loss. They are therefore considering
closing this store. However, before doing so, they have sought your advice as an external business consultant.

Required:
Advise the directors of Book Co on whether they should close Store A, based upon the predicted loss.

Answer
Before a decision is made on closing Store A, the budgeted income statement should be reformatted to clearly
identify the relevant costs, ie the costs that would be avoided if the store were closed. This has been done in Table
10.2.

Table 10.2

Store A Store B Store C Total


$000 $000 $000 $000
Sales 1,000 1,800 1,200 4,000
Less variable costs:
Cost of sales 480 720 510 1,710
Store expenses 25 40 22 87
Warehouse costs (40%) 48 80 56 184
Contribution to all fixed costs 447 960 612 2,019
Store specific fixed costs:
Store staff salaries 140 180 150 470
Store running costs 65 95 70 230
Advertising 60 60 60 180
Contribution to central fixed costs 182 625 332 1,139
Central fixed costs:
Head office expenses 540
Warehouse costs (60%) 276
Net profit/(loss) 323

The original budgeted income statement suggests that Store A will make a loss of $25,000. Therefore, if the store
were closed, the company would expect overall profit to increase by $25,000 to $348,000.
However, the reformatted budget demonstrates that Store A is making a positive contribution of $182,000 towards
the common fixed costs. This means that if Store A were closed, overall company profit would fall by $182,000 to only
$141,000.
The conclusion should be, therefore, that because Store A is making a positive contribution to common fixed costs,
it should remain open. If Store A is closed, this will have a significant negative impact on overall company profit.

Expert view 10.5: Allocated and apportioned fixed costs


In Chapter 8 we looked at the concept of allocating and apportioning fixed costs when establishing the total
cost of a department or product. When decision making, it is important to be aware of the distinction between
allocated and apportioned fixed costs, because this distinction usually has an impact upon whether the costs
are relevant. In Worked Example 10.8 the allocated fixed costs of running Store A were relevant to the decision
because they would be avoidable if the store were closed. On the other hand, the apportioned head office fixed
costs were not relevant as these costs would still be incurred even if Store A were closed.

Relevant costing: summary


In this section covering relevant costs in different decision-making situations, we have covered
the following important points:

When decision making, only relevant costs should be considered. Any costs deemed not to
be relevant should be ignored.
Relevant costs are those future costs that will be changed by a particular decision. This may
include opportunity costs.
Whether any given cost is relevant will depend upon the situation.
The time horizon chosen will impact upon what costs become relevant for a given situation.

Conclusion
In this chapter we have covered two important financial decision-making techniques: CVP
analysis and relevant costing. CVP analysis enables managers to identify the levels of operating
activity that are necessary to avoid losses and achieve targeted levels of profits. It also enables
managers to analyse the impact of organizational changes and the related risks. Relevant costing
is a useful technique of focusing management attention on those costs that will be affected by
decisions.

COMPREHENSION QUESTIONS
1. Explain the concept of ‘margin of safety’ in CVP analysis.
2. If a business reduces its fixed costs whilst maintaining the same contribution margin, will this increase or decrease
its break-even point?
3. Explain what is meant by ‘operating gearing’.
4. If a company increases its operating gearing, will this make its profit more or less sensitive to changes in sales
volume?
5. Explain three potential weaknesses of CVP analysis as a decision tool.
6. Alpha Co has 200 units of material x in inventory. This originally cost $40 per unit. Alpha Co no longer needs the
material x for its original use. It could sell the inventory for $10 per unit. However, the production manager wishes
to use the material x in a new project. What is the relevant cost of using the 200 units of material x in the new
project?
7. Give an example of a cost which may not be relevant in the short term, but could become relevant in a long-term
decision.

Answers on pages 403–405.

Exercises
Answers on pages 405–408.

Exercise 10.1: Target profit


Huath Co has identified that it can optimize its production costs if it sells 200,000 units of its product. The
directors wish to make a profit of $300,000. The following costs have been predicted:

Direct material cost $15 per unit


Direct labour cost $10 per unit
Variable production overhead cost $20 per unit
Fixed costs $400,000 per annum

Calculate the sales price per unit which the company needs to apply.

Exercise 10.2: CVP analysis and price change


Beath Co makes and sells mobile phone batteries. The variable costs of production are $4 and the current
sales price is $7. Fixed costs are $35,000 per month and the annual profit of the company is currently
$270,000. The volume of sales demand is constant throughout the year.
The company is considering lowering the sales price to $6 to stimulate sales, but is uncertain of the effect on
sales volume.

(a) Calculate the minimum volume of sales required to justify the reduction in price.
(b) What would the percentage change in profit be if sales increased to 300,000 units?

Exercise 10.3: CVP analysis and selling price


Luis Co is a manufacturer of precision parts for electric motors. The company has developed a new brush
assembly for large industrial electric motors. The company expects to sell 20,000 units of this new product in
the following year, and wishes to make a profit of $90,000. Costs are as follows:

Direct material cost $25 per unit


Direct labour cost $10 per unit
Variable production overhead cost $25 per unit
Fixed costs $70,000 per annum

Required:
(a) Calculate the required sales price per unit.
(b) Calculate the break-even sales price.
Exercise 10.4: Make or buy?
Nuin Co manufactures a number of different components for the oil industry. The management is considering
whether to buy in or continue making one of the components (component A). This component currently has a
manufacturing cost as follows:

$
Direct labour (4hr @ $12ph) 48
Direct materials 24
Variable overheads (4hr @ $2ph) 8
Fixed overheads (4hr @ $5ph) 20
100 per unit

The direct labour, direct materials and variable overhead costs all relate directly to the production of component
A and would not be incurred if production of the component stopped. However, the fixed overheads charge is
an apportionment of costs which would still be incurred even if component A were not produced.

Required:
Under each of the three (separate) situations below, advise the management of Nuin Co whether component A
should be bought in or made in-house:

(a) The purchasing manager has found an external manufacturer that can supply the component A at a
guaranteed price of $90 per unit.
(b) The external supplier can offer component A at $90 per unit. If Nuin Co continues to manufacture
component A in-house, it will need to install new computer-controlled manufacturing systems which will
have a fixed cost of $50,000 per year.
(C) The external supplier can offer component A at $90 per unit. The manufacture of component A in-house
requires the use of specialist skilled direct labour. If component A were bought in, that direct labour could
be used in the production of component B which is sold for $180 and has a manufacturing cost as follows:

$
Direct labour (8hr @ $12ph) 96
Direct materials 18
Variable overheads (8hr @ $2ph) 16
Fixed overheads (8hr @ $5ph) 40
170 per unit

Answers to comprehension questions


1. The margin of safety is the gap, measured in units of sales, between the expected level of sales and the
break-even point, ie it is the volume of sales above the break-even point. It is therefore a useful measure
of how far sales can fall before the business starts to make a loss.
2. The break-even point in the units is measured as:

Therefore, if fixed costs are reduced but the contribution margin remains the same, the break-even point
will decrease.
3. Operating gearing describes the relationship between an organization’s fixed costs and variable costs:

4. If a company increases its operating gearing, it increases fixed costs as a proportion of total costs. This
means that total costs will change less with changes in volume of sales. If the volume of sales increases,
profitability will increase faster. However, if the volume of sales decreases, profitability will decrease
faster.
5. CVP analysis has a number of weaknesses because of the way it simplifies the complexities of the real
world:
(a) CVP analysis assumes that unit price remains the same as the volume of sales changes.
(b) CVP analysis assumes that fixed costs remain constant across all volumes of activity.
(c) CVP analysis assumes that variable costs per unit remain constant across all volumes of activity.
(d) CVP analysis assumes that costs can be clearly categorized into either variable or fixed.
(e) CVP analysis typically focuses on one accounting period. It therefore ignores longer-term strategic
issues.
(f) CVP analysis usually assumes either a single product or a constant mix of products.
(g) CVP analysis assumes a simple, single-stage manufacturing process.
(h) CVP analysis assumes that the forces influencing a business are static rather than dynamic.
6. The original cost of $40 per unit is a sunk cost and is therefore not relevant. If Alpha Co uses material x
in the new project, there will be an opportunity cost of $10 per unit, which is the lost opportunity of
earning that revenue from selling the inventory. The relevant cost is therefore $2,000.
7. The only costs that are relevant to a decision are those that will change as a result of the decision.
Therefore, any cost that will not change is not a relevant cost. There are many costs which any
organization may be unable to avoid in the short term and which would therefore be incurred no matter
what decision is made, but which could be reduced or avoided in the long term. One example is the
rental on a building. If there is a contractual agreement to pay the rent, this will continue to be a cost even
if the business decides that it will no longer use the building. However, in the longer term the business
could sell the lease of the building or let it lapse, thereby removing the cost.

Answers to exercises

Exercise 10.1: Target profit

Huath Co should charge a price of $55 per unit.

Exercise 10.2: CVP analysis and price change


(a) The minimum volume of sales required to justify the reduction in price: Beath Co currently earns an annual
profit of $270,000 by selling batteries at $7 each. In order to justify a reduction in the sales price the
company must make at least the same level of profit at the new price of $6.

(b) The percentage change in profit if sales increased to 300,000 units: New profit at sales of 300,000 units:

Exercise 10.3: CVP analysis and selling price


(a) The required sales price per unit: The sales price must earn sufficient contribution to cover fixed costs and
earn the target profit of $95,000. We should therefore use the formula:

(b) The break-even sales price:


Exercise 10.4: Make or buy?
(a) The relevant production costs for this decision are those costs which would be avoided if in-house
production of component A were to stop. These are:

$
Direct labour (4hr @ $12ph) 48
Direct materials 24
Variable overheads (4hr @ $2ph) 8
80 per unit

Therefore the relevant cost of producing component A is $80 per unit, compared to the relevant cost of
buying in which is $90 per unit. The company should therefore continue to make component A in-house.
(b) If Nuin Co continues to manufacture component A in-house, it will incur an additional fixed cost of $50,000
per year. It will still be worth manufacturing in-house if the volume of production is sufficient to cover the
extra cost. (This therefore becomes an exercise in CVP analysis.)
As per the calculations in part (a) above, there is a marginal benefit of $10 per unit in manufacturing
component A in-house ($90 – $80). The break-even point for this benefit will be:

Therefore, if Nuin Co expects to sell more than 5,000 units per year of component A, it will be financially
better to continue producing in-house. However, if the company expects to sell fewer than 5,000 units per
year, it would be better off by buying in component A.
(c) If the skilled direct labour could be used elsewhere within the business, there is an opportunity cost in using
it to produce component A. This opportunity cost will be measured in terms of the lost contribution from
component B resulting from using the direct labour to produce component A.
The contribution per unit from component B is:

Sales price – Variable production costs = $180 – ($96 + $18 + $16) = $50

However, each unit of component B requires eight hours of direct labour, whereas each unit of component
A requires only four hours of direct labour. Therefore, each unit of component A produced will take away
50 per cent of the contribution from component B, ie $25.
The relevant cost of producing component A is therefore:

$
Direct labour (4hr @ $12 p.h.) 48
Direct materials 24
Variable overheads (4hr @ $2 p.h.) 8
Opportunity cost 25
105 per unit

It would therefore be financially better for Nuin Co to buy in component A at $90 and to use the direct
labour to produce component B.
Appendix A
An introduction to double-entry
bookkeeping

Double-entry bookkeeping (bookkeeping, for short) is best taught through t-accounts; named
thus because of their shape. T-accounts have a left side and a right side. Hereafter these will
be referred to as the DEBIT side and the CREDIT side. This links back to the golden rule of
accounting: that every debit must have an equal and opposite credit otherwise your ledgers
would not balance.
Over the next few pages we will walk you through each of the Mobius Inc examples,
completing the t-accounts as we go.
Bookkeeping comes naturally to a small number of people, but to the majority, however, this
is something which needs to be practised and given careful consideration. Regardless of whether
you ‘get it’ within five minutes or five years, for almost everyone there is a ‘light-bulb moment’
where everything clicks into place and you reflect uncomprehendingly on the times you couldn’t
see how it worked.

Approaching double-entry bookkeeping

Step 1: Set up your t-accounts


A t-account captures the key information about a transaction:

the date;
an outline description of the transaction;
– note that this is normally the name of the account the other side of the journal entry is
going to;
the value.

For example, our first transaction involves two t-accounts: cash at bank and in hand; and share
capital. The cash at bank and in hand t-account would appear as shown in Table A.1.

Table A.1 Cash at bank and in hand


Note that there the ‘$’ or ‘value’ columns need to be totalled. That is because each t-account
needs to ‘balance out’ at the end of the period of account:

On the one hand, statement of financial position balances, for example cash at bank and in
hand don’t stop existing at the end of a period and therefore the amounts left over can be
carried forward to the next period.
On the other hand, income statement accounts, for example rental costs, relate to a period of
account and once they have been incurred and paid, they should be closed out.
– More on this later…

Step 2: Write out your journal entry


Certainly, while you’re in the initial learning stages of bookkeeping it is good practice to write
out every journal entry. As you progress, you might stop doing this. Note, however, that if you
don’t write out the entry in full and you make an error (eg a transposition error, post two debits
instead of a debit and credit, miss one side of the transaction), tracing the error back is more
difficult.
The question states:

Mobius Inc (1)


On day 1, you opt to put financial distance between you and the trading entity and transfer
$1,000 from your personal bank account to a bank account you hold in the name of the new
enterprise – Mobius Inc.
The journal entry is as follows:

$ $
Debit Cash at bank and in hand 1,000
Credit Share capital: equity and reserves 1,000

Useful tip
People who struggle with double-entry bookkeeping often don’t understand the relevance of
posting an entry as either a debit or a credit. The interconnectedness of position and performance
provides the clue that we need. The grid below might help you to understand.
First, however, remember the accounting equation:
Assets MINUS Liabilities EQUALS Equity + Reserves
(Assets – Liabilities = Equity and Reserves)

For this equation to work, an increase in assets would necessarily lead to either:

a decrease in another asset;


an increase in liabilities; or
an increase in equity and reserves.

Examples might be:

a decrease in another asset, eg swapping cash for inventory;


an increase in liabilities, eg buying inventory on credit from a supplier;
an increase in equity and reserves, eg selling share capital for cash.

To continue…
We know that the profit is taken to equity and reserves to close it out at the end of a period. A
simple double entry effecting both the statement of financial position and income statement (ie
performance and position) might be:

An operating expense being paid, eg rent.


Cash would reduce (decrease in assets) and the rental cost in the income statement would
increase.
– Assuming this was the only transaction a business undertook during the period of account,
at the end of the year this rental expense would be the loss for the period. This balance
would be taken to retained earnings (a sub-heading under reserves).
– Therefore, our accounting equation holds: a decrease in assets is offset by a decrease in
equity and reserves.

The accounting equation tells us that there are two equal and opposite sides to every transaction.
This can be summarized as shown in Table A.2.

Table A.2

Debit Credit
Assets increase Assets decrease
Assets and liabilities (excluding equity Liabilities decrease Liabilities increase
and reserves) (ie positive effect on financial (ie negative effect on financial
position) position)
Expenses increase Expenses decrease
Income statement (Revenue decreases) (Revenue increases)
and (ie negative impact on performance; (ie positive impact on performance;
Equity and reserves reduce profit) increase profit)
Equity and reserves decrease Equity and reserves increase
Or, in shorthand (Table A.3).

Table A.3

Debit Credit
Assets and liabilities (excluding equity and reserves) + –
Income statement and Equity and reserves – +

Step 3: Post this journal entry to the relevant t-accounts


(Table A.4)
Table A.4

Cash at bank and in hand


Date Description $ $ Description Date
Day 1 Initial investment capital 1,000
Share capital
Date Description $ $ Description Date
1,000 Bank Day 1

Step 4: Close out the t-accounts


Assuming that this is the only transaction, you now need to move your closing balances to a trial
balance and then into the financial statements (Table A.5).

Table A.5

Cash at bank and in hand


Date Description $ $ Description Date
Beginning of day 1 Balance brought forward –
Day 1 Initial investment capital 1,000
1,000 Balance carried forward End of day 1
1,000 1,000
Beginning of day 2 Balance brought forward 1,000
Share capital
Date Description $ $ Description Date
– Balance brought forward Beginning of day 1
1,000 Bank Day 1
End of day 1 Balance carried forward 1,000
1,000 1,000
1,000 Balance brought forward Beginning of day 2

Let’s walk through what has happened here.


The original double entry was posted. Both sides were posted correctly and to the appropriate
t-accounts.
If this were the only transaction of the day, we need to ask ourselves whether these two
balances are position or performance related, ie were they assets or liabilities, or were they
revenue or expenses. It is obvious that both cash and share capital are position-related balances.
They will continue to be controlled (obligated to settle) by the entity on day 2 given that nothing
happened to cancel them out on day 1.
Therefore, these balances have been carried forward (c/fwd) from day 1 to day 2.
You will notice that when the account is closed out, the closing balance which we carry
forward is the balancing entry to make the account sum:

Mobius Inc has $1,000 in the bank at the end of day 1 and none was spent and none
generated during that 24-hour period. Therefore, the company has $1,000 at the beginning
of day 2.
Mobius Inc has $1,000 of share capital. No more was sold and none was repurchased during
day 1, therefore they close with $1,000 and open the following day with $1,000.

When you open up the account, be careful to bring the brought forward amount down on the
correct side, ie the same as it was accumulated; or the opposite of the balancing entry that closed
it out. In this case, the cash is c/fwd as a credit to balance the account out, but b/fwd as a debit.
Reflecting on the grid above, you’ll see that a debit signifies a positive balance and, indeed, we
hold $1,000 in the bank. If you had brought this balance forward as a credit, this would mean
that you opened with a negative amount of cash (bank overdraft).

Tutor’s note: The joy of learning accounting is that it is happening all the time, both to you, with you and all around
you. Next time you go into a shop, think about the transaction as a bookkeeping exercise.

Do you have cash and you want to exchange it for a new pen? If so, you exchange one asset
for another (credit cash [reduce asset]; credit equipment [increase asset]).
Do you pay for your electricity by direct debit? If so, you exchange an asset (credit bank
[reduce asset]; debit expenses [reduce profits]).

Step 5: Summarize into a ‘trial balance’


A trial balance (TB) is a summary of your t-accounts’ closing balances, ie a list of closing
balances. Our TB is simple in this instance as we have only two accounts/balances.
Your TB needs to have three columns:

1. account name, eg cash at bank and in hand;


2. debit;
3. credit.

Both debit and credit columns are then summed. If they are not the same value, you have a TB
that does not balance and you can therefore conclude that you’ve made a mistake in your
bookkeeping somewhere.
Our TB at the end of day 1 should appear as shown in Table A.6.
Table A.6

Debit (Dr) Credit (Cr)


$ $
Cash at bank and in hand 1,000
Share capital 1,000
1,000 1,000

NOTE: You will often see debit and credit shortened to ‘dr’ (DR, Dr) and ‘cr’ (CR, Cr).

Step 6: Extract the information from the TB and draw up


your financial statements
Again in this case, the exercise is straightforward. We only have the statement of financial
position entries. We have an asset and the equity (share capital). The statement would appear as
follows:

Statement of financial position


For Mobius Inc
As at the end of Day 1
Assets
Current Assets
Cash at bank and in hand 1,000
Total Assets 1,000
Equity and reserves
Share capital 1,000
Total 1,000

Tutor’s Note: These are the basics of double-entry bookkeeping. We have walked through this example slowly and
carefully, explaining points of common error as we go. We will now move up a gear and proceed through the rest of
the examples at a slightly speedier pace. If you struggle, however, then come back to these basic points.

Mobius Inc

Day 2
The text states:

On day 2, you borrow $500 from a friend to provide further financial help to your business.

The journal entry would be as follows:

$ $
Debit Cash at bank and in hand 500
Credit Loan (liabilities) 500
The t-accounts would read as shown in Table A.7.

Table A.7

Cash at bank and in hand


Date Description $ $ Description Date
Beginning of day 1 Balance brought forward –
Day 1 Initial investment capital 1,000
1,000 Balance carried forward End of day 1
1,000 1,000
Beginning of day 2 Balance brought forward 1,000
Day 2 Loan 500
1,500 Balance carried forward End of day 2
1,500 1,500
Beginning of day 3 Balance brought forward 1,500
Loan account (liabilities)
Date Description $ $ Description Date
– Balance brought forward Beginning of day 2
500 Bank Day 2
End of day 2 Balance carried forward 500
500 500
500 Balance brought forward Beginning of day 3
Share capital
Date Description $ $ Description Date
– Balance brought forward Beginning of day 1
1,000 Bank Day 1
End of day 1 Balance carried forward 1,000
1,000 1,000
1,000 Balance brought forward Beginning of day 2
End of day 2 Balance carried forward 1,000
1,000 1,000
1,000 Balance brought forward Beginning of day 3

In other words, no change on the share capital account from day 1.


The revised TB at the end of day 2 should read as shown in Table A.8.

Table A.8

Day 2 Debit Credit


(Dr) (Cr)
$ $
Cash at bank and in hand 1,500
Loan 500
Share capital 1,000
1,500 1,500

Thus, the statement of financial position as at the end of day 2 should read as follows:
Statement of financial position
For Mobius Inc
As at the end of Day 2
Assets
Current assets
Cash at bank and in hand 1,500
Liabilities
Non-current liabilities
Loan (500)
Net assets 1,000
Equity and reserves
Share capital 1,000
Total 1,000

Days 3 & 4
On day 3, Mobius Inc invests $500 of cash by acquiring a new computer.

$ $
Debit Non-current assets (computer) 500
Credit Cash 500

On day 4, Mobius Inc buys some raw materials worth $400 and holds them as inventories. The
cash required to settle the invoices related to these purchases does not need to be found for 10
days as these are the credit terms offered.

$ $
Debit Inventories 400
Credit Trade payables 400

We now have six t-accounts open:

1. cash at bank and in hand;


2. non-current assets;
3. inventories;
4. trade payables;
5. share capital (remain unchanged during days 3 & 4);
6. loan account (remain unchanged during days 3 & 4).

And they should appear as shown in Table A.9.

Table A.9

Cash at bank and in hand


Date Description $ $ Description Date
Beg of day 3 Bal. b/fwd 1,500
500 Inventories Day 3
1,000 Bal. c/fwd End of day 4
1,500 1,500
Non-current assets (computer)
Date Description $ $ Description Date
Beg of day 3 Bal. b/fwd –
Day 3 Bank 500
500 Bal. c/fwd End of day 4
500 500
Inventories (raw materials)
Date Description $ $ Description Date
Beg of day 3 Bal. b/fwd –
Day 4 Trade payables 400
400 Bal. c/fwd End of day 4
400 400
Trade payables
Date Description $ $ Description Date
– Bal. b/fwd Beg of day 3
400 Inventories Day 4
End of day 4 Bal. c/fwd 400
400 400
Share capital
Date Description $ $ Description Date
1,000 Bal. b/fwd Beg of day 3
End of day 4 Bal. c/fwd 1,000
1,000 1,000
Loan
Date Description $ $ Description Date
500 Bal. b/fwd Beg of day 3
End of day 4 Bal. c/fwd 500
500 500

The statement of financial position as at the end of day 4 should appear as follows:

Statement of financial position


For Mobius Inc
As at the end of Day 4
Assets
Non-current assets
Computer 500
Current assets
Inventories (raw materials) 400
Cash at bank and in hand 1,000
Liabilities
Non-current liabilities
Loan (500)
Current liabilities
Trade payables (400)
Net assets 1,000
Equity and reserves
Share capital 1,000
Total 1,000

Days 5 & 6
Days 5 & 6 introduce performance-related transactions into our accounting. This means that we
will need to produce a statement of financial position and an income statement. The bookkeeping
remains exactly the same as before.

NOTE: In the interests of economy and ease of reading, we will no longer produce the t-accounts where there have
been no changes during the period (in this case, days 5 & 6 eg share capital). We strongly advise that whilst you are
in the learning phase, however, you continue to produce all the accounts and get into the habit of opening them up
and closing them down.

On day 5, Mobius Inc uses the raw materials to produce 30 units of finished goods stock.
On day 6, half of these are sold for $50 per unit. Cash is received immediately for five of
those sold. The remainder were sold to customers on 10-day credit terms.
The first part of this transaction (day 5) is straightforward. We are simply moving an asset
from one pot to another. In this case, raw materials are converted into finished goods.

$ $
Debit Inventories: finished goods 400
Credit Inventories: raw materials 400

The second part, however, is a little more tricky as explained in the main text. We advise you do
this in two steps: firstly, the revenue side of the transaction; and secondly, the cost of the
transaction (cost of sales), as follows:

$ $
Debit Cash 250
Debit Trade receivables 500
Credit Revenue 750

And then:

Debit Cost of sales 200


Credit Inventories: finished goods 200

See Table A.10.

Table A.10

Inventories (raw materials)


Date Description $ $ Description Date
Beg of day 4 Bal. b/fwd 400
400 Transfer to finished goods Day 5
– Bal. c/fwd End of day 6
400 400
Inventories (finished goods)
Date Description $ $ Description Date
Beg of day 4 Bal. b/fwd –
Day 5 Transferred from raw materials 400 Transfer to finished goods Day 5
200 Cost of sales Day 6
200 Bal. c/fwd End of day 6
400 400
Cash at bank and in hand
Date Description $ $ Description Date
Beg of day 5 Bal. b/fwd 1,000
Day 6 Revenue 250
1,250 Bal. c/fwd End of day 4
1,250 1,250
Trade receivables
Date Description $ $ Description Date
Beg of day 5 Bal. b/fwd –
Day 6 Revenue 500
500 Bal. c/fwd End of day 4
500 500

Note that these are all statement of financial position accounts and all hold either assets or
liabilities which will be carried forward to the next period of account. This is not true of income
statement accounts, however. These will be closed out in the period and any balance taken to the
income statement (and then through to reserves to be carried forward as a net balance at the end
of the period).
The income statement accounts are as shown in Table A.11.

Table A.11

Revenue
Date Description $ $ Description Date
250 Cash Day 6
500 Trade receivables Day 6
Take to income statement 750
750 750
Cost of sales
Date Description $ $ Description Date
Day 6 Inventories 200
200 Take to income statement
200 200

The financial statements would appear as follows:


Days 8 to 14
Note: The financial statements are produced in the body of the text and therefore shall not be
duplicated here. See Table A.12.

Table A.12

Inventories (raw materials)


Date Description $ $ Description Date
Beg of day 8 Bal. b/fwd –
Day 8 Trade payables 800
800 Transfer to finished goods Day 9
– Bal. c/fwd End of day 14
800 800
Trade payables
Date Description $ $ Description Date
400 Bal. b/fwd Beg of day 8
800 Inventories (raw materials) Day 8
Day 14 Bank 400
End of day 14 Bal. c/fwd 800
1,200 1,200
Inventories (finished goods)
Date Description $ $ Description Date
Beg of day 8 Bal. b/fwd 200
Day 9 Transferred from raw materials 800
467 Cost of sales
533 Bal. c/fwd End of day 14
1,000 1,000
Operating expenses
Date Description $ $ Description Date
Day 10 Bank (utilities) 200
Day 12 Bank (stationery) 50
Day 14 Bank (web development costs) 750
1,000 Take to income statement
1,000 1,000
Cash at bank and in hand
Date Description $ $ Description Date
Beg of day 8 Bal. b/fwd 1,250
200 Utilities Day 10
Day 11 Revenue 1,200
100 Scanner Day 12
50 Stationery Day 12
Day 13 Trade receivables 500
400 Trade payables Day 14
750 Web development costs Day 14
1,450 Bal. c/fwd End of day 14
2,950 2,950
Trade receivables
Date Description $ $ Description Date
Beg of day 8 Bal. b/fwd 500
Day 11 Revenue 900
500 Bank Day 13
900 Bal. c/fwd End of day 14
1,400 1,400
Revenue
Date Description $ $ Description Date
1,200 Cash Day 11
900 Trade receivables Day 11
Take to income statement 1,800
1,800 2,100
Cost of sales
Date Description $ $ Description Date
Day 11 Inventories 467
467 Take to income statement
467 467
Non-current assets (scanner)
Date Description $ $ Description Date
Beg of day 8 Bal. b/fwd –
Day 12 Bank 600
600 Bal. c/fwd End of day 14
600 600

The trial balance as at the end of day 14 should now read as shown in Table A.13.

Table A.13

Day 14 Debit Credit


(Dr) (Cr)
$ $
Non-current assets (computer) 500
Non-current assets (scanner) 100
Inventories (raw materials) –
Inventories (finished goods) 533
Trade receivables 900
Cash at bank and in hand 1,450
Trade payables 800
Loan 500
Share capital 1,000
Retained earnings (b/fwd from week 1) 550
Revenue 2,100
Cost of sales 467
Operating expenses 1,000
4,950 4,950

Through to the end of month 1


Mobius (5): Adjustments for non-current assets
The following journal entries relate to the depreciation of the non-current assets which the
company acquired during the first two weeks, ie the scanner and the computer:

Depreciation charge against the scanner


Debit Depreciation (income statement) 5
Credit Computer: Accumulated depreciation (statement of financial position) 5
Depreciation charge against the computer
Debit Depreciation (income statement) 10
Credit Scanner: Accumulated depreciation (statement of financial position) 10

Note that the t-accounts for the accumulated depreciation are kept separately to the t-accounts for
the original cost of the asset. This is so that later adjustments required – for example, disposal or
re-measurement – are simpler.
The entries related to the acquisition of the motor vehicle, the associated loan required to buy
the asset and the unpaid interest on that loan would be as follows:

Acquisition of motor vehicle


Debit Motor vehicle: cost 20,000
Debit Motor vehicle costs (additional extras) 3,350
Credit Loan (to acquire vehicle) 2,350
Depreciation charge against the motor vehicle
Debit Depreciation (income statement) 154
Credit Motor Vehicle: Accumulated depreciation (statement of financial position) 154
Interest on loan (unpaid) used to acquire the motor vehicle
Debit Finance costs 90
Credit Interest accrual 90

Mobius (6): Period-end adjustments


Telephone line installation unpaid as at the end of the month
Debit Operating expenses: Telephone line installation 100
Credit Accruals 100
Rent paid in advance (prepayment account required)
Debit Prepayment (rent) 1,000
Credit Bank 1,000
Sales and related receivables account
Debit Trade receivables 20,000
Credit Revenue 20,000
Debit Bank 16,000
Credit Trade receivables 16,000
Adjustments for inventories, trade payables and cost of sales
Debit Inventories (raw materials) 9,000
Credit Trade payables 9,000
Debit Inventories (finished goods) 8,000
Credit Inventories (raw materials) 8,000
Debit Cost of sales 6,000
Credit Inventories (finished goods) 6,000
Debit Cost of sales 533
Credit Inventories (finished goods) 533
Debit Trade payables 800
Debit Trade payables 7,000
Credit Bank 7,800
Appendix B
International Accounting/Financial
Reporting Standards (as at 02.04.2013)

NOTE: This list excludes the work of the Interpretations Committee.

International Financial Reporting Standards


# Name Issued
IFRS 1 First-time Adoption of International Financial Standards 2008*
IFRS 2 Share-based Payment 2004
IFRS 3 Business Combinations 2008*
IFRS 4 Insurance Contracts 2004
IFRS 5 Non-current Assets Held for Sale and Discontinued Operations 2004
IFRS 6 Exploration for and Evaluation of Mineral Assets 2004
IFRS 7 Financial Instruments: Disclosures 2005
IFRS 8 Operating Segments 2006
IFRS 9 Financial Instruments 2010*
IFRS 10 Consolidated Financial Statements 2011
IFRS 11 Joint Arrangements 2011
IFRS 12 Disclosure of Interests in Other Entities 2011
IFRS 13 Fair Value Measurement 2011

International Accounting Standards


# Name Issued
IAS 1 Presentation of Financial Statements 2007*
IAS 2 Inventories 2005*
Consolidated Financial Statements
IAS 3 1976
Superseded in 1989 by IAS 27 and IAS 28
Depreciation Accounting
IAS 4
Withdrawn in 1999
Information to Be Disclosed in Financial Statements
IAS 5 1976
Superseded by IAS 1 effective 1 July 1998
Accounting Responses to Changing Prices
IAS 6
Superseded by IAS 15, which was withdrawn December 2003
IAS 7 Statement of Cash Flows 1992
IAS 8 Accounting Policies, Changes in Accounting Estimates and Errors 2003
Accounting for Research and Development Activities
IAS 9 Superseded by IAS 39 effective 1 July 1999
IAS 10 Events After the Reporting Period 2003
IAS 11 Construction Contracts 1993
IAS 12 Income Taxes 1996*
Presentation of Current Assets and Current Liabilities
IAS 13
Superseded by IAS 39 effective 1 July 1998
Segment Reporting
IAS 14 1997
Superseded by IFRS 8 effective 1 January 2009
Information Reflecting the Effects of Changing Prices
IAS 15 2003
Withdrawn December 2003
IAS 16 Property, Plant and Equipment 2003*
IAS 17 Leases 2003*
IAS 18 Revenue 1993*
Employee Benefits
IAS 19 1998
Superseded by IAS 19 (2011) effective 1 January 2013
IAS 19 Employee Benefits (2011) 2011*
IAS 20 Accounting for Government Grants and Disclosure of Government Assistance 1983
IAS 21 The Effects of Changes in Foreign Exchange Rates 2003*
Business Combinations
IAS 22 1998*
Superseded by IFRS 3 effective 31 March 2004
IAS 23 Borrowing Costs 2007*
IAS 24 Related Party Disclosures 2009*
Accounting for Investments
IAS 25
Superseded by IAS 39 and IAS 40 effective 2001
IAS 26 Accounting and Reporting by Retirement Benefit Plans 1987
IAS 27 Separate Financial Statements (2011) 2011
Consolidated and Separate Financial Statements
IAS 27 2003
Superseded by IFRS 10, IFRS 12 and IAS 27 (2011) effective 1 January 2013
IAS 28 Investments in Associates and Joint Ventures (2011) 2011
Investments in Associates
IAS 28 2003
Superseded by IAS 28 (2011) and IFRS 12 effective 1 January 2013
IAS 29 Financial Reporting in Hyperinflationary Economies 1989
Disclosures in the Financial Statements of Banks and Similar Financial Institutions
IAS 30 1990
Superseded by IFRS 7 effective 1 January 2007
Interests In Joint Ventures
IAS 31 2003*
Superseded by IFRS 11 and IFRS 12 effective 1 January 2013
IAS 32 Financial Instruments: Presentation 2003*
IAS 33 Earnings Per Share 2003*
IAS 34 Interim Financial Reporting 1998
Discontinuing Operations
IAS 35 1998
Superseded by IFRS 5 effective 1 January 2005
IAS 36 Impairment of Assets 2004*
IAS 37 Provisions, Contingent Liabilities and Contingent Assets 1998
IAS 38 Intangible Assets 2004*
Financial Instruments: Recognition and Measurement
IAS 39 2003*
Superseded by IFRS 9 effective 1 January 2015
IAS 40 Investment Property 2003*
IAS 41 Agriculture 2001
Other pronouncements
Name Issued
Conceptual Framework for Financial Statements 2010 2010
Preface to International Financial Reporting Standards 2002
IFRS for Small and Medium Sized Entities 2009
IFRS Practice Statement Management Commentary 2010
APPENDIX C
Example earnings announcements

IBM plc 1Q13 Press Release


Diluted EPS:
– GAAP: $2.70, up 3 per cent;
– Operating (non-GAAP): $3.00, up 8 per cent;
Net income:
– GAAP: $3.0 billion, down 1 per cent;
– Operating (non-GAAP): $3.4 billion, up 3 per cent;
Gross profit margin:
– GAAP: 45.6 per cent, up 0.6 points;
– Operating (non-GAAP): 46.7 per cent, up 1.0 points;
Revenue: $23.4 billion, down 5 per cent, down 3 per cent adjusting for currency;
Free cash flow of $1.7 billion, down $0.2 billion;
Software revenue flat, up 1 per cent adjusting for currency;
– Pre-tax: income up 4 per cent; margin up 1.2 points;
Services revenue down 4 per cent, down 1 per cent adjusting for currency;
– Pre-tax: income up 10 per cent; margin up 2.0 points;
Services backlog of $141 billion, up 1 per cent, up 5 per cent adjusting for currency;
– Closed 22 deals of more than $100 million in the quarter;
Systems and Technology revenue down 17 per cent, down 16 per cent adjusting for
currency;
Growth markets revenue down 1 per cent, up 1 per cent adjusting for currency;
Business analytics revenue up 7 per cent;
Smarter Planet revenue up more than 25 per cent;
Cloud revenue up more than 70 per cent;
Reiterating full-year 2013 operating (non-GAAP) EPS expectation of at least $16.70.

ARMONK, N.Y., April 18, 2013 … IBM (NYSE: IBM) today announced first-quarter 2013
diluted earnings of $2.70 per share, a year-to-year increase of 3 per cent. Operating (non-GAAP)
diluted earnings were $3.00 per share, compared with operating diluted earnings of $2.78 per
share in the first quarter of 2012, an increase of 8 per cent.
First-quarter net income was $3.0 billion, down 1 per cent year-to-year. Operating (non-
GAAP) net income was $3.4 billion compared with $3.3 billion in the first quarter of 2012, an
increase of 3 per cent.
Total revenues for the first quarter of 2013 of $23.4 billion were down 5 per cent (down 3 per
cent, adjusting for currency) from the first quarter of 2012.
“In the first quarter, we grew operating net income, earnings per share and expanded
operating margins but we did not achieve all of our goals in the period. Despite a solid start and
good client demand we did not close a number of software and mainframe transactions that have
moved into the second quarter. The services business performed as expected with strong profit
growth and significant new business in the quarter,” said Ginni Rometty, IBM chairman,
president and chief executive officer.
“Looking ahead, in addition to closing those transactions, we expect to benefit from
investments we are making in our growth initiatives and from the actions we are taking to
improve under-performing parts of the business. We remain confident in this model of
continuous transformation and in our ability to deliver our full-year 2013 operating earnings per
share expectation of at least $16.70.”

First-Quarter GAAP – Operating (non-GAAP)


Reconciliation
First-quarter operating (non-GAAP) diluted earnings exclude $0.30 per share of charges: $0.12
per share for the amortization of purchased intangible assets and other acquisition-related
charges, and $0.18 per share for retirement-related charges.

Full-Year 2013 Expectations


IBM is reiterating its expectation for full-year 2013 GAAP diluted earnings per share of at least
$15.53. Operating (non-GAAP) diluted earnings per share expectations remain at least $16.70.
The 2013 operating (non-GAAP) earnings expectations exclude $1.17 per share of charges for
amortization of purchased intangible assets, other acquisition-related charges, and retirement-
related charges.

Geographic Regions
The Americas’ first-quarter revenues were $10.0 billion, a decrease of 4 per cent (down 3 per
cent, adjusting for currency) from the 2012 period. Revenues from Europe/Middle East/Africa
were $7.3 billion, down 4 per cent (down 4 per cent, adjusting for currency). Asia-Pacific
revenues decreased 7 per cent (down 1 per cent, adjusting for currency) to $5.7 billion. OEM
revenues were $426 million, down 16 per cent compared with the 2012 first quarter.

Growth Markets
Revenues from the company’s growth markets decreased 1 per cent (up 1 per cent, adjusting for
currency). Revenues in the BRIC countries – Brazil, Russia, India and China – decreased 1 per
cent (up 3 per cent, adjusting for currency).

Services
Global Technology Services segment revenues decreased 4 per cent (down 2 per cent, adjusting
for currency) to $9.6 billion. Global Business Services segment revenues were down 3 per cent
(flat, adjusting for currency) to $4.5 billion.
Pre-tax income from Global Technology Services was up 7 per cent and pre-tax margin
increased to 16.1 per cent. Global Business Services pre-tax income increased 17 per cent and
pre-tax margin increased to 15.1 per cent.
The estimated services backlog at March 31 was $141 billion, up 1 per cent year over year at
actual rates (up 5 per cent, adjusting for currency). The company closed 22 service agreements of
more than $100 million in the quarter.

Software
Revenues from the Software segment were flat at $5.6 billion (up 1 per cent, adjusting for
currency) compared with the first quarter of 2012. Software pre-tax income increased 4 per cent
and pre-tax margin increased to 31.5 per cent.
Revenues from IBM’s key middleware products, which include WebSphere, Information
Management, Tivoli, Social Workforce Solutions (formerly Lotus) and Rational products, were
$3.5 billion, up 1 per cent (up 2 per cent, adjusting for currency) versus the first quarter of 2012.
Operating systems revenues of $578 million were down 2 per cent (down 1 per cent, adjusting
for currency) compared with the prior-year quarter.
Revenues from the WebSphere family of software products increased 6 per cent year over
year. Information Management software revenues decreased 2 per cent. Revenues from Tivoli
software increased 1 per cent. Revenues from Social Workforce Solutions (formerly Lotus)
software increased 8 per cent, and Rational software decreased 2 per cent.

Hardware
Revenues from the Systems and Technology segment totalled $3.1 billion for the quarter, down
17 per cent (down 16 per cent, adjusting for currency) from the first quarter of 2012. Excluding
Retail Store Solutions (RSS), revenues were down 14 per cent (down 13 per cent, adjusting for
currency). Systems and Technology pre-tax loss increased $0.3 billion.
Total systems revenues, excluding RSS, decreased 13 per cent (down 13 per cent, adjusting
for currency). Revenues from System z mainframe server products increased 7 per cent
compared with the year-ago period. Total delivery of System z computing power, as measured in
MIPS (millions of instructions per second), increased 27 per cent. Revenues from Power
Systems were down 32 per cent compared with the 2012 period. Revenues from System x were
down 9 per cent. Revenues from System Storage decreased 11 per cent. Revenues from
Microelectronics OEM decreased 16 per cent.

Financing
Global Financing segment revenues were up 2 per cent (up 4 per cent, adjusting for currency) in
the first quarter at $499 million. Pre-tax income for the segment increased 5 per cent to $538
million.

Gross Profit
The company’s total gross profit margin was 45.6 per cent in the 2013 first quarter compared
with 45.1 per cent in the 2012 first-quarter period. Total operating (non-GAAP) gross profit
margin was 46.7 per cent in the 2013 first quarter compared with 45.7 per cent in the 2012 first-
quarter period, with increases in Global Technology Services and Global Business Services.

Expense
Total expense and other income decreased 3 per cent to $7.1 billion, compared to the prior-year
period. S,G&A expense of $5.6 billion decreased 5 per cent year over year. R,D&E expense of
$1.6 billion increased 3 per cent, compared with the year-ago period. Intellectual property and
custom development income decreased to $183 million compared with $255 million a year ago.
Other (income) and expense was income of $60 million compared with prior-year income of $58
million. Interest expense decreased to $94 million compared with $110 million in the prior year.
Total operating (non-GAAP) expense and other income decreased 4 per cent to $6.9 billion
compared with the prior-year period. Operating (non-GAAP) S,G&A expense of $5.4 billion
decreased 7 per cent compared with prior-year expense. Operating (non-GAAP) R,D&E expense
of $1.6 billion increased 1 per cent compared with the year-ago period.

***

Pre-tax income decreased 6 per cent to $3.6 billion. Pre-tax margin decreased 0.1 points to 15.4
per cent. Operating (non-GAAP) pre-tax income decreased 1 per cent to $4.1 billion and pre-tax
margin was 17.4 per cent, up 0.8 points.
IBM’s tax rate was 15.9 per cent, down 4.1 points year over year; operating (non-GAAP) tax
rate was 17.3 per cent, down 3.2 points compared to the year-ago period. The lower tax rate is
primarily due to benefits recorded to reflect changes in tax laws enacted during the quarter,
including the reinstatement of the US Research and Development Tax Credit.
Net income margin increased 0.5 points to 13.0 per cent. Total operating (non-GAAP) net
income margin increased 1.2 points to 14.4 per cent.
The weighted-average number of diluted common shares outstanding in the first-quarter 2013
was 1.12 billion compared with 1.17 billion shares in the same period of 2012. As of March 31,
2013, there were 1.11 billion basic common shares outstanding.
Debt, including Global Financing, totalled $33.4 billion, compared with $33.3 billion at year-
end 2012. From a management segment view, Global Financing debt totalled $25.2 billion
versus $24.5 billion at year-end 2012, resulting in a debt-to-equity ratio of 7.2 to 1. Non-global
financing debt totalled $8.2 billion, a decrease of $0.6 billion since year-end 2012, resulting in a
debt-to-capitalization ratio of 34.3 per cent from 36.1 per cent.
IBM ended the first-quarter 2013 with $12.0 billion of cash on hand and generated free cash
flow of $1.7 billion, excluding Global Financing receivables, down approximately $0.2 billion
year over year. The company returned $3.5 billion to shareholders through $0.9 billion in
dividends and $2.6 billion of gross share repurchases. The balance sheet remains strong, and the
company is well positioned to support the business over the long term.

Forward-Looking and Cautionary Statements


Except for the historical information and discussions contained herein, statements contained in
this release may constitute forward-looking statements within the meaning of the Private
Securities Litigation Reform Act of 1995. Forward-looking statements are based on the
company’s current assumptions regarding future business and financial performance. These
statements involve a number of risks, uncertainties and other factors that could cause actual
results to differ materially, including the following: a downturn in economic environment and
corporate IT spending budgets; the company’s failure to meet growth and productivity
objectives, a failure of the company’s innovation initiatives; risks from investing in growth
opportunities; failure of the company’s intellectual property portfolio to prevent competitive
offerings and the failure of the company to obtain necessary licences; cybersecurity and data
privacy considerations; fluctuations in financial results and purchases, impact of local legal,
economic, political and health conditions; adverse effects from environmental matters, tax
matters and the company’s pension plans; ineffective internal controls; the company’s use of
accounting estimates; the company’s ability to attract and retain key personnel and its reliance on
critical skills; impacts of relationships with critical suppliers and business with government
clients; currency fluctuations and customer financing risks; impact of changes in market liquidity
conditions and customer credit risk on receivables; reliance on third party distribution channels;
the company’s ability to successfully manage acquisitions and alliances; risk factors related to
IBM securities; and other risks, uncertainties and factors discussed in the company’s Form 10-Q,
Form 10-K and in the company’s other filings with the US Securities and Exchange Commission
(SEC) or in materials incorporated there in by reference. Any forward-looking statement in this
release speaks only as of the date on which it is made. The company assumes no obligation to
update or revise any forward-looking statements.

Presentation of Information in this Press Release


In an effort to provide investors with additional information regarding the company’s results as
determined by generally accepted accounting principles (GAAP), the company has also disclosed
in this press release the following non-GAAP information which management believes provides
useful information to investors:
IBM results and expectations –

presenting operating (non-GAAP) earnings per share amounts and related income statement
items;
presenting non-global financing debt-to-capitalization ratio;
adjusting for free cash flow;
adjusting for currency (ie, at constant currency);
adjusting for the divestiture of RSS.

The rationale for management’s use of non-GAAP measures is included as part of the
supplemental materials presented within the first-quarter earnings materials.These materials are
available via a link on the IBM investor relations website at www.ibm.com/investor and are
being included in Attachment II (“Non-GAAP Supplemental Materials”) to the Form 8-K that
includes this press release and is being submitted today to the SEC.

Conference Call and Webcast


IBM’s regular quarterly earnings conference call is scheduled to begin at 4:30 pm EDT, today.
The webcast may be accessed via a link at http://www.ibm.com/investor/vents/1q13.phtml.
Presentation charts will be available shortly before the webcast.
Financial Results Below (certain amounts may not add due to use of rounded numbers;
percentages presented are calculated from the underlying whole-dollar amounts).

Cal Chalco 2012


INTERNATIONAL BUSINESS MACHINES CORPORATION COMPARATIVE
FINANCIAL RESULTS (Unaudited; Dollars in millions except per share amounts)
INTERNATIONAL BUSINESS MACHINES CORPORATION CONSOLIDATED
STATEMENT OF FINANCIAL POSITION (Unaudited)
INTERNATIONAL BUSINESS MACHINES CORPORATION CASH FLOW ANALYSIS
(Unaudited)
INTERNATIONAL BUSINESS MACHINES CORPORATION SEGMENT DATA
(Unaudited)
INTERNATIONAL BUSINESS MACHINES CORPORATION US GAAP TO
OPERATING RESULTS RECONCILIATION (Unaudited; Dollars in millions except
per share amounts)
Hong Kong Exchanges and Clearing Limited and The Stock Exchange of Hong Kong Limited
take no responsibility for the contents of this announcement, make no representation as to its
accuracy or completeness and expressly disclaim any liability whatsoever for any loss
howsoever arising from or in reliance upon the whole or any part of the contents of this
announcement.
This announcement is made by Aluminum Corporation of China Limited* (the “Company” or,
together with its subsidiaries, the “Group”) pursuant to Part XIVA of the Securities and Futures
Ordinance (Chapter 571 of the Laws of Hong Kong) and Rules 13.09 and 13.10B of the Rules
Governing the Listing of Securities on The Stock Exchange of Hong Kong Limited.

1 Important Notice
1.1 board of directors, the supervisory committee, the directors, supervisors and senior
management of the Company guarantee that this report contains no false representation,
misleading statement or material omission. All of them jointly and severally accept responsibility
for the truthfulness, accuracy and completeness of the contents of this report.
1.2 All the directors of the Company attended the Board meeting.
1.3 The first quarterly financial report of the Company has not been audited.
1.4 Name of Person-in-charge of the Company Xiong Weiping
Name of Person-in-charge of Accounting Xie Weizhi
Name of Head of the Accounting Department Lu Dongliang

Xiong Weiping, Person-in-charge of the Company, Xie Weizhi, Person-in-charge of Accounting,


and Lu Dongliang, Head of the Accounting Department warrant the truthfulness and
completeness of the financial statements in this quarterly report.

2 Company Profile
2.1 Principal accounting information and financial indicators
Deducting the gains and losses arising from extraordinary items and
amount:
2.2 Total number of shareholders and the top 10 shareholders not subject to
trading moratorium as at the end of the reporting period
Unit: Share
Total number of shareholders as at the end of the reporting period 526,111

The top 10 shareholders of tradable shares not subject to trading moratorium


3 Significant Events
3.1 Material changes in major accounting items and financial indicators and
the reasons thereof

1. Business tax and surcharges increased by 75 per cent, mainly attributable to the rising
turnover tax deriving from the improved net income or expense from other operations of the
Group.
2. Financial expenses increased by 41 per cent, mainly attributable to the larger size of
interest-bearing debts from last year resulting from the new incorporation of the
consolidated statements of China Aluminum Ningxia Energy Group Co., Ltd. into the
Group.
3. Loss on asset impairment increased by 8,589 per cent, mainly attributable to the increase in
the provision for inventory impairment following the lower price of major products of the
Group at the end of the period.
4. Loss on fair value changes increased by 464 per cent, mainly attributable to the increasing
floating loss on fair value changes arising from futures contracts held by the Group.
5. Investment income increased by 80 per cent, mainly attributable to the increase in gains
from associated companies and joint ventures resulting from the new incorporation of the
consolidated statements of China Aluminum Ningxia Energy Group Co., Ltd. into the
Group.
6. Non-operating income increased by 1,289 per cent, mainly attributable to the higher fair
value of the mining rights over the considerations engendered as a result of acquiring China
Aluminum Ningxia Energy Group Co., Ltd. by the Group.
7. Non-operating expenses decreased by 43 per cent, mainly attributable to less donations
made by the Group.
8. Total profit increased by 33 per cent, mainly attributable to the enhanced gross profit
margin resulting from the strict control on costs, fees and expenses as well as the expansion
of income sources.
9. Income tax expenses increased by 89 per cent, mainly attributable to the non-provision of
deferred income tax assets for most of the losses occurred during the period.
10. Minority interests increased by 177 per cent, mainly attributable to the profit increase of
subsidiaries controlled by the Group.
11. Held-for-trading financial assets decreased by 80 per cent, mainly attributable to less
floating profit on the positions of the futures and forward foreign exchange contracts of the
Group.
12. Notes receivables increased by 46 per cent, mainly attributable to the tightening national
currency policy, which increased the use of bank’s acceptance notes in settlement.
13. Accounts receivables increased by 219 per cent, mainly attributable to the extension of the
credit period as a result of the tightening national currency policy, keen market competition
and the increase in trade volume, as well as the new incorporation of the consolidated
statements of China Aluminum Ningxia Energy Group Co., Ltd.
14. Prepayments increased by 65 per cent, mainly attributable to the prepayments for certain
procurement with a view to expanding trading channels.
15. Interest receivable increased by 2,834 per cent, mainly attributable to the increasing accrued
interests resulted from the entrusted loans to associates.
16. Other receivables increased by 40 per cent, mainly attributable to the additional entrusted
loans to associates.
17. Other current assets increased by 52 per cent, mainly attributable to the new incorporation
of the consolidated statements of China Aluminum Ningxia Energy Group Co., Ltd. into the
Group, which increased deductible input tax under value-added tax.
18. Investment properties increased from nil at the beginning of the period, mainly attributable
to the new incorporation of the consolidated statements of China Aluminum Ningxia
Energy Group Co., Ltd. into the Group.
19. Construction materials increased by 119 per cent, mainly attributable to new incorporation
of the consolidated statements of China Aluminum Ningxia Energy Group Co., Ltd.
20. Intangible assets increased by 156 per cent, mainly attributable to the new incorporation of
the consolidated statements of China Aluminum Ningxia Energy Group Co., Ltd. into the
Group, which increased raising the number of coal mining rights.
21. Financial liabilities held for trading increased by 288 per cent, mainly attributable to the
increased floating loss on the positions of the futures and forward foreign exchange
contracts of the Group.
22. Notes payable increased by 186 per cent, mainly attributable to the tightening of national
currency policy, resulting in a proper increase of notes used in settling procurement.
23. Accounts payables increased by 46 per cent, mainly attributable to the tightening of national
currency policy, resulting in the proper extension of credit period in procurement.
24. Advances from customers increased by 74 per cent, mainly attributable to the portion of
payment received in advance based on clients’ credibility.
25. Accrued interest increased by 77 per cent, mainly attributable to the larger size of interest-
bearing debts of the Group.
26. Dividends payable increased by 71 per cent, mainly attributable to the cash dividends
payable accounted by subsidiaries under the Group.
27. Long-term loans increased by 85 per cent, mainly attributable to the new incorporation of
consolidated statements of China Aluminum Ningxia Energy Group Co., Ltd. into the
Group, making additional long-term bank loans.
28. Special payables increased by 108 per cent, mainly attributable to the government subsidies
added by the new incorporation of consolidated statements of China Aluminum Ningxia
Energy Group Co., Ltd. into the Group.
29. Projected liabilities increased from nil at the beginning of the period, mainly attributable to
the projected mining rights considerations payable upon acquisition of China Aluminum
Ningxia Energy Group Co., Ltd.
30. Other non-current liabilities increased by 58 per cent, mainly attributable to the additional
deferred income arising from the new incorporation of consolidated statements of China
Aluminum Ningxia Energy Group Co., Ltd. into the Group.
31. Minority interests increased by 39 per cent, mainly attributable to minority interests added
by the new incorporation of consolidated statements of China Aluminum Ningxia Energy
Group Co., Ltd. into the Group.
32. For the first quarter of 2013, the Group recorded a revenue of RMB34.213 billion. Included
in its total profit was a loss of RMB948 million, representing a reduction of loss of
RMB466 million or 32.96 per cent from the loss of RMB1.414 billion for the corresponding
period last year. Vested in the net profit attributable to owners of the parent company was a
loss of RMB975 million, representing a reduction of loss of RMB113 million or 10.42 per
cent from the loss of RMB1.088 billion for the corresponding period last year.
33. For the first quarter of 2013, the Group had increased its operating gross profit by 37 per
cent compared to the corresponding period last year, of which, cost reduction of major
products ranged between 6 per cent and 9 per cent, cutting the loss by approximately
RMB1.4 billion. However, price reduction of major products ranged between 5 per cent and
7 per cent compared to the corresponding period last year, reducing the profit by
approximately RMB1.3 billion.
34. For the first quarter of 2013, the Group continued to increase its effort in controlling the
production and operating costs. During the period, selling expenses amounted to RMB458
million, representing an increase of RMB12 million from RMB446 million for the same
period last year. Upon deduction of the additional RMB14 million resulting from the
incorporation of China Aluminum Ningxia Energy Group Co., Ltd., such expenses
decreased slightly as compared with the same period last year. Administrative expenses for
the period amounted to RMB645 million, representing an increase of RMB23 million as
compared with RMB622 million for the same period last year. Upon deduction of additional
RMB80 million resulting from the incorporation of China Aluminum Ningxia Energy
Group Co., Ltd., such expenses decreased by RMB57 million as compared with the same
period last year. Finance expenses for the period amounted to RMB1.518 billion,
representing an increase of RMB442 million as compared with RMB1.076 billion for the
same period of last year. Upon deduction of additional RMB277 million resulting from the
incorporation of China Aluminum Ningxia Energy Group Co., Ltd., such expenses
increased by RMB165 million as compared with the same period last year, mainly
attributable to the increase in the size of interest-bearing debts of the Group, offset by a
decrease of 0.43 percentage point in the weighted average interest rate in the end of the first
quarter of 2013 as compared with the end of the first quarter of last year.

3.2 Progress of significant events and effects thereof and analysis on


solutions

3.3 Implementation of undertakings by the Company, its shareholders and


de facto controller

1. During the A share issue of the Company in 2007, Aluminum Corporation of China
(“Chinalco”) undertook that Chinalco would arrange to dispose of its aluminum fabrication
business, or the Company would acquire the aluminum fabrication business from Chinalco,
and acquire the pseudo-boehmite business from Chinalco within a certain period of time
following the listing of the Company’s A shares.
In 2008, the Company successfully bid for the five aluminum fabrication enterprises
under the control of Chinalco in an open tender process through the equity exchange. Since
the market conditions for pseudo-boehmite are immature, Chinalco does not propose to
inject its pseudo-boehmite business to the Company’s portfolio. When conditions become
mature, Chinalco will continue to duly complete the matters undertaken within the time
limit.
2. On 22 August 2011, the Company issued a letter of undertaking of Aluminum Corporation
of China Limited* to resolve the horizontal competition with Jiaozuo Wanfang Aluminum
Company Limited in the aluminum business (
), pursuant to which it undertook to make its best endeavours to eliminate by
proper means the competition in aluminum business with Jiaozuo Wanfang within five
years.

3.4 Warning on any potential loss in accumulated net profit for the period
from the beginning of the year to the end of the next reporting period or any
material change from the corresponding period last year and the reason
thereof

3.5 Implementation of cash dividend policy during the reporting period


The terms for the distribution of cash dividend were prescribed in the Articles of Association of
the Company: 1) the Company takes full account of the returns to investors and distributes
dividends to shareholders every year according to the prescribed ratio of distributable profit for
the prevailing year; 2) the Company upkeeps a continuous and stable profit distribution policy
whilst considering the long-term interest of the Company, the overall interest of its shareholders
as a whole and the sustainable development of the Company; 3) cash dividends shall be
considered first when the Company distributes dividends.

No final dividend was distributed during the reporting period.


Aluminum Corporation of China Limited*
Legal representative: Xiong Weiping
26 April 2013

4 Appendices
4.1 Consolidated Balance Sheet
As at 31 March 2013
Prepared by: Aluminum Corporation of China Limited*

Unit: Thousand RMB, Unaudited


Balance Sheet of the Parent Company
As at 31 March 2013
Prepared by: Aluminum Corporation of China Limited*

Unit: Thousand RMB, Unaudited


4.2 Consolidated Income Statement
January to March 2013
Prepared by: Aluminum Corporation of China Limited*

Unit: Thousand RMB, Unaudited


Income Statement of the Parent Company
January to March 2013
Prepared by: Aluminum Corporation of China Limited*

Unit: Thousand RMB, Unaudited


4.3 Consolidated Cash Flow Statement
January to March 2013
Prepared by: Aluminum Corporation of China Limited*

Unit: Thousand RMB, Unaudited


Cash Flow Statement of the Parent Company
January to March 2013
Prepared by: Aluminum Corporation of China Limited*

Unit: Thousand RMB, Unaudited


As at the date of this announcement, the members of the Board of Directors comprise Mr. Xiong
Weiping, Mr. Luo Jianchuan and Mr. Liu Xiangmin (Executive Directors); Mr. Shi Chungui and
Mr. Lv Youqing and Mr. Liu Caiming (Non-executive Directors); Mr. Zhang Zhuoyuan, Mr.
Wang Mengkui and Mr. Zhu Demiao (Independent Non-executive Directors).

* For identification purpose only


Appendix D
Discount tables

Table Appendix D.1

Years 1% 2% 3% 4% 5% 6% 7% 8% 9% 10%
1 0.990 0.980 0.971 0.962 0.952 0.943 0.935 0.926 0.917 0.909
2 0.980 0.961 0.943 0.925 0.907 0.890 0.873 0.857 0.842 0.826
3 0.971 0.942 0.915 0.889 0.864 0.840 0.816 0.794 0.772 0.751
4 0.961 0.924 0.888 0.855 0.823 0.792 0.763 0.735 0.708 0.683
5 0.951 0.906 0.863 0.822 0.784 0.747 0.713 0.681 0.650 0.621
6 0.942 0.888 0.837 0.790 0.746 0.705 0.666 0.630 0.596 0.564
7 0.933 0.871 0.813 0.760 0.711 0.665 0.623 0.583 0.547 0.513
8 0.923 0.853 0.789 0.731 0.677 0.627 0.582 0.540 0.502 0.467
9 0.914 0.837 0.766 0.703 0.645 0.592 0.544 0.500 0.460 0.424
10 0.905 0.820 0.744 0.676 0.614 0.558 0.508 0.463 0.422 0.386
11 0.896 0.804 0.722 0.650 0.585 0.527 0.475 0.429 0.388 0.350
12 0.887 0.788 0.701 0.625 0.557 0.497 0.444 0.397 0.356 0.319
13 0.879 0.773 0.681 0.601 0.530 0.469 0.415 0.368 0.326 0.290
14 0.870 0.758 0.661 0.577 0.505 0.442 0.388 0.340 0.299 0.263
15 0.861 0.743 0.642 0.555 0.481 0.417 0.362 0.315 0.275 0.239
16 0.853 0.728 0.623 0.534 0.458 0.394 0.339 0.292 0.252 0.218
17 0.844 0.714 0.605 0.513 0.436 0.371 0.317 0.270 0.231 0.198
18 0.836 0.700 0.587 0.494 0.416 0.350 0.296 0.250 0.212 0.180
19 0.828 0.686 0.570 0.475 0.396 0.331 0.277 0.232 0.194 0.164
20 0.820 0.673 0.554 0.456 0.377 0.312 0.258 0.215 0.178 0.149
Years 11% 12% 13% 14% 15% 16% 17% 18% 19% 20%
1 0.901 0.893 0.885 0.877 0.870 0.862 0.855 0.847 0.840 0.833
2 0.812 0.797 0.783 0.769 0.756 0.743 0.731 0.718 0.706 0.694
3 0.731 0.712 0.693 0.675 0.658 0.641 0.624 0.609 0.593 0.579
4 0.659 0.636 0.613 0.592 0.572 0.552 0.534 0.516 0.499 0.482
5 0.593 0.567 0.543 0.519 0.497 0.476 0.456 0.437 0.419 0.402
6 0.535 0.507 0.480 0.456 0.432 0.410 0.390 0.370 0.352 0.335
7 0.482 0.452 0.425 0.400 0.376 0.354 0.333 0.314 0.296 0.279
8 0.434 0.404 0.376 0.351 0.327 0.305 0.285 0.266 0.249 0.233
9 0.391 0.361 0.333 0.308 0.284 0.263 0.243 0.225 0.209 0.194
10 0.352 0.322 0.295 0.270 0.247 0.227 0.208 0.191 0.176 0.162
11 0.317 0.287 0.261 0.237 0.215 0.195 0.178 0.162 0.148 0.135
12 0.286 0.257 0.231 0.208 0.187 0.168 0.152 0.137 0.124 0.112
13 0.258 0.229 0.204 0.182 0.163 0.145 0.130 0.116 0.104 0.093
14 0.232 0.205 0.181 0.160 0.141 0.125 0.111 0.099 0.088 0.078
15 0.209 0.183 0.160 0.140 0.123 0.108 0.095 0.084 0.074 0.065
16 0.188 0.163 0.141 0.123 0.107 0.093 0.081 0.071 0.062 0.054
17 0.170 0.146 0.125 0.108 0.093 0.080 0.069 0.060 0.052 0.045
18 0.153 0.130 0.111 0.095 0.081 0.069 0.059 0.051 0.044 0.038
19 0.138 0.116 0.098 0.083 0.070 0.060 0.051 0.043 0.037 0.031
20 0.124 0.104 0.087 0.073 0.061 0.051 0.043 0.037 0.031 0.026
Appendix E
Annuity factors

Table Appendix E.1

Years 1% 2% 3% 4% 5% 6% 7% 8% 9% 10%
1 0.990 0.980 0.971 0.962 0.952 0.943 0.935 0.926 0.917 0.909
2 1.970 1.942 1.913 1.886 1.859 1.833 1.808 1.783 1.759 1.736
3 2.941 2.884 2.829 2.775 2.723 2.673 2.624 2.577 2.531 2.487
4 3.902 3.808 3.717 3.630 3.546 3.465 3.387 3.312 3.240 3.170
5 4.853 4.713 4.580 4.452 4.329 4.212 4.100 3.993 3.890 3.791
6 5.795 5.601 5.417 5.242 5.076 4.917 4.767 4.623 4.486 4.355
7 6.728 6.472 6.230 6.002 5.786 5.582 5.389 5.206 5.033 4.868
8 7.652 7.325 7.020 6.733 6.463 6.210 5.971 5.747 5.535 5.335
9 8.566 8.162 7.786 7.435 7.108 6.802 6.515 6.247 5.995 5.759
10 9.471 8.983 8.530 8.111 7.722 7.360 7.024 6.710 6.418 6.145
11 10.368 9.787 9.253 8.760 8.306 7.887 7.499 7.139 6.805 6.495
12 11.255 10.575 9.954 9.385 8.863 8.384 7.943 7.536 7.161 6.814
13 12.134 11.348 10.635 9.986 9.394 8.853 8.358 7.904 7.487 7.103
14 13.004 12.106 11.296 10.563 9.899 9.295 8.745 8.244 7.786 7.367
15 13.865 12.849 11.938 11.118 10.380 9.712 9.108 8.559 8.061 7.606
16 14.718 13.578 12.561 11.652 10.838 10.106 9.447 8.851 8.313 7.824
17 15.562 14.292 13.166 12.166 11.274 10.477 9.763 9.122 8.544 8.022
18 16.398 14.992 13.754 12.659 11.690 10.828 10.059 9.372 8.756 8.201
19 17.226 15.678 14.324 13.134 12.085 11.158 10.336 9.604 8.950 8.365
20 18.046 16.351 14.877 13.590 12.462 11.470 10.594 9.818 9.129 8.514
Years 11% 12% 13% 14% 15% 16% 17% 18% 19% 20%
1 0.901 0.893 0.885 0.877 0.870 0.862 0.855 0.847 0.840 0.833
2 1.713 1.690 1.668 1.647 1.626 1.605 1.585 1.566 1.547 1.528
3 2.444 2.402 2.361 2.322 2.283 2.246 2.210 2.174 2.140 2.106
4 3.102 3.037 2.974 2.914 2.855 2.798 2.743 2.690 2.639 2.589
5 3.696 3.605 3.517 3.433 3.352 3.274 3.199 3.127 3.058 2.991
6 4.231 4.111 3.998 3.889 3.784 3.685 3.589 3.498 3.410 3.326
7 4.712 4.564 4.423 4.288 4.160 4.039 3.922 3.812 3.706 3.605
8 5.146 4.968 4.799 4.639 4.487 4.344 4.207 4.078 3.954 3.837
9 5.537 5.328 5.132 4.946 4.772 4.607 4.451 4.303 4.163 4.031
10 5.889 5.650 5.426 5.216 5.019 4.833 4.659 4.494 4.339 4.192
11 6.207 5.938 5.687 5.453 5.234 5.029 4.836 4.656 4.486 4.327
12 6.492 6.194 5.918 5.660 5.421 5.197 4.988 4.793 4.611 4.439
13 6.750 6.424 6.122 5.842 5.583 5.342 5.118 4.910 4.715 4.533
14 6.982 6.628 6.302 6.002 5.724 5.468 5.229 5.008 4.802 4.611
15 7.191 6.811 6.462 6.142 5.847 5.575 5.324 5.092 4.876 4.675
16 7.379 6.974 6.604 6.265 5.954 5.668 5.405 5.162 4.938 4.730
17 7.549 7.120 6.729 6.373 6.047 5.749 5.475 5.222 4.990 4.775
18 7.702 7.250 6.840 6.467 6.128 5.818 5.534 5.273 5.033 4.812
19 7.839 7.366 6.938 6.550 6.198 5.877 5.584 5.316 5.070 4.843
20 7.963 7.469 7.025 6.623 6.259 5.929 5.628 5.353 5.101 4.870
References

Chapter 1
International Federation of Accountants (2011) A Proposed Definition of ‘Professional Accountant’, Agenda Paper 7-B
Jones, M (2011) Creative Accounting, Fraud and International Accounting Scandals, John Wiley & Sons Ltd, Chichester, UK
Kaplan, R S (2011) Accounting scholarship that advances professional knowledge and practice, The Accounting Review, 86 (2),
pp 367–83
Pacioli, L (1497) Summa de arithmetica, geometria, proportioni et proportionalita, Venice, Italy

Supplementary reading
Carnegie, G D and Napier, C J (2010) Traditional accountants and business professionals: portraying the accounting profession
after Enron, Accounting, Organizations and Society, 35 (3), pp 360–76
DeCoster, D T and Rhode, J G (1971) The accountant’s stereotype: real or imagined, deserved or unwarranted, Accounting
Review, 46, pp 651–64
International Accounting Standards Board (2010) The Conceptual Framework for Financial Reporting 2010, IFRS Foundation
Publications Department, London

Chapter 2
Gallagher, M and Paul, B (2012) Assessing Going Concern: Stakeholders would benefit from clarity in US disclosure
requirements, Point of view, PricewaterhouseCooper LLP Publications, December

Supplementary reading
Bakar, N B A and Said, J M (2007) Historical cost versus current cost accounting, Accountants Today, January
Elliott, B and Elliott, J (2011) Financial Accounting and Reporting, 15th edn, Pearson Education, Harlow
Fahnestock, R T and Bostwick, E D (2011) An analysis of the fair value controversy, Proceedings of The American Society of
Business and Behavioral Sciences at Las Vegas, 18 (1), pp 910–21
Hendriksen, E F and Van Breda, M F (1992) Accounting Theory, 5th edn, Irwin Professional Publishing, Burr Ridge, IL
Institute of Chartered Accountants in England and Wales (2009) Going concern: don’t panic, Accountancy, January
Laux, C and Leuz, C (2009) The crisis of fair-value accounting: making sense of the recent debate, Accounting, Organizations
and Society, 34 (6), pp 826–34
Power, M (2010) Fair value accounting, financial economics and the transformation of reliability, Accounting and Business
Research, 40 (3), pp 197–220

Chapter 3
Chen, K H and Shimerda, T A (1981) An empirical analysis of useful financial ratios, Financial Management, pp 51–60
Ryanair Holdings plc (accessed 18 December 2013) Annual report 2012 [Online] http://www.ryanair.com/en/investor/download/
Chapter 4
Abraham, S, Marston, C and Darby, P (2012) Risk Reporting: Clarity, relevance and location, Institute of Chartered Accountants
of Scotland, Research Report, ISBN 978-1-904574-87-3
Cyert, R M and March, J G (1963) A Behavioral Theory of the Firm, Prentice-Hall, Englewood Cliffs, NJ
Frankel, R, Johnson, M and Skinner, D J (1999) An empirical examination of conference calls as a voluntary disclosure medium,
Journal of Accounting Research, 37 (1), pp 133–50
Mitchell, R K, Agle, B R and Wood, D J (1997) Toward a theory of stakeholder identification and salience: defining the principle
of who and what really counts, Academy of Management Review, 22 (4), pp 853–86

Supplementary reading
Barker, R G (1998) The market for information – evidence from finance directors, analysts and fund managers, Accounting and
Business Research, 29 (1), pp 3–20
Barker, R, Hendry, J, Roberts, J, et al (2012) Can company-fund manager meetings convey informational benefits? Exploring the
rationalisation of equity investment decision making by UK fund managers, Accounting, Organizations and Society, 37 (4), pp
207–22
Brown, S, Hillegeist, S A and Lo, K (2004) Conference calls and information asymmetry, Journal of Accounting and Economics,
37 (3), pp 343–66
Cornelissen, J (2011) Corporate Communication: A guide to theory and practice, 3rd edn, Sage, London
Frankel, R, Johnson, M and Skinner, D J (1999) An empirical examination of conference calls as a voluntary disclosure medium,
Journal of Accounting Research, 37 (1), pp 133–50
Mayew, W and Venkatachalam, M (2012) The power of voice: managerial affective states and future firm performance, Journal
of Finance, 67 (1), pp 1–44
Mayew, W J, Sharp, N Y and Venkatachalam, M (2013) Using earnings conference calls to identify analysts with superior private
information, Review of Accounting Studies, 18, pp 386–413

Chapter 5
Comshare, Inc (2000) Comshare Survey of Top Financial Executives: Planning and budgeting today [Online]
www.comshare.com
Libby, T and Lindsay, R M (2010) Beyond budgeting or budgeting reconsidered? A survey of North-American budgeting
practice, Management Accounting Research, 21 (1), pp 56–75
Pyhrr, P (1973) Zero-base Budgeting: A practical management tool for evaluating expenses, John Wiley & Sons, Inc, New York

Supplementary reading
Hansen, S C (2004) The Closed Loop: Implementing activity-based planning and budgeting, Bookman, Indianapolis, IN
Hansen, S C, Otley, D T and Van der Stede, W A (2003) Practice developments in budgeting: an overview and research
perspective, Journal of Management Accounting Research, 15, pp 95–116

Chapter 6
Johnson, H T and Kaplan, R S (1987) Relevance Lost: The rise and fall of management accounting, Harvard Business School
Press, Boston, MA
Kaplan, R S and Norton, D P (1992) The balanced scorecard – measures that drive performance, Harvard Business Review, 70
(1), pp 71–79
Kaplan, R S and Norton, D P (1996) The Balanced Scorecard: Translating strategy into action, Harvard Business School Press,
Boston, MA
Kaplan, R S and Norton, D P (2008) The Execution Premium: Linking strategy to operations for competitive advantage, Harvard
Business School Press, Boston, MA
Starovic, D (2003) Performance Reporting to Boards: A guide to good practice, CIMA, London

Chapter 7

Supplementary reading
Baumol, W J (1952) The transactions demand for cash: an inventory theoretic approach, Quarterly Journal of Economics, 66, pp
545–56
Miller, M H and Orr, D (1966) A model of the demand for money by firms, Quarterly Journal of Economics, 80 (3), pp 413–35
Tobin, J (1956) The interest elasticity of the transactions demand for cash, Review of Economics and Statistics, 38 (3), pp 241–47

Chapter 8
McCarthy, J E (1960) Basic Marketing: A managerial approach, Richard D Irwin, Homewood, IL
Schulz, D E, Tannenbaum, S I and Lauterborn, R F (1993) Integrated Marketing Communications, NTC Publishing Group,
Lincolnwood, IL
Shimizu, K (1973) Advertising Theory and Strategies, Souseisha Book Company, Japan

Chapter 9

Supplementary reading
Ogier, T, Rugman, J and Spicer, L (2004) Real Cost of Capital: A business field guide to better financial decisions, Financial
Times/Prentice Hall, Harlow
Smit, H T J and Trigeorgis, L (2004) Strategic Investment: Real options and games, Princeton University Press, Princeton, NJ

Chapter 10
Peters, T J and Waterman, R H (1982) In Search of Excellence: Lessons from America’s best-run companies, Profile Books,
London
Index

NB: page numbers in italic indicate figures or tables

ABC inventory management (i), (ii)


Abraham, Santhosh, Marston, Claire and Darby, Phil (i)
absorption costing (i)
accountancy, introduction to (i)
accounting research (i)
business, types of (i)
business purpose (i)
limited company (i), (ii)
partnership (i)
sole trader (i)
comprehension questions (i)
answers (i)
definitions
accountant (i)
accounting (i)
exercises (i)
answers (i)
financial accountants (i), (ii), (iii)
Big Four firms (i)
function on (i)
qualifications (i)
services provided by (i)
users of (i)
financial statements, overview (i)
fit together, how they (i), (ii)
problems, key (i)
required information (i)
terminology (i)
Tesco example (i), (ii), (iii), (iv), (v), (vi)
information, useful (i), (ii)
enhancing characteristics (i)
fundamental characteristics (i)
materiality (i)
management accounting (i), (ii)
origins of (i)
regulation (i), (ii), (iii)
Conceptual Framework (F) (i), (ii)
IFRS Foundation (i), (ii)
IFRS Interpretations Committee (i)
International Accounting Standards Board (IASB) (i), (ii)
principles (i)
accounting rate of return (ARR) (i)
AccountingWEB UK (i)
accounts payable, managing (i)
accounts receivable, managing (i)
credit setting and control (i), (ii)
debt collection (i)
debt factoring (i)
invoice discounting (i)
accruals concept (i)
acid test ratio (i)
activity-based budgets (i), (ii)
activity-based costing (ABC) and pricing (i)
alternative depreciation methods (i)
Aluminium Corporation of China Limited (CHALCO) (i)
Amazon (i), (ii)
Anglo American (i)
annuities (i)
Apple (i)
iPad (i)
stratified pricing (i)
Asda (i)
asset financing (i)
finance lease (i)
hire purchase (HP) (i)
operating lease (i)
Association of Chartered Certified Accountants (ACCA) (i)

balanced scorecard (i), (ii)


Balanced Scorecard, The (i)
Baumol–Tobin model (i), (ii), (iii)
worked example (i)
Bernstein, Michelle (i)
Beyond Budgeting Round Table (BBRT) (i), (ii)
BHP Billiton (i)
Big Four firms (i)
BM plc (i)
BMW Group plc (i)
Bonderman, David (i)
Bradford & Bingley (i)
break-even point (BEP) (i), (ii)
budgeting (i)
activity-based budgets (i), (ii)
’budget’, definition of (i)
comprehension questions (i)
answers (i)
exercises (i)
answers (i)
importance of (i)
limitations of (i)
behaviour, influence on (i)
cost (i)
volatile environments (i)
merchandizing businesses (i)
not-for-profits (i)
planning and control (i), (ii)
preparing (i)
sequence of events (i)
worked example (i)
production businesses (i)
public sector businesses (i)
rolling budgets (i)
service businesses (i)
setting (i), (ii), (iii)
bottom-up (i)
financial budget (i)
incremental (i)
master budget (i)
operational budget (i)
top-down (i), (ii)
zero-based budgets (i)
bundle pricing (i)
business entity concept (i)
business, types of (i)
business purpose (i)
limited company (i), (ii)
partnership (i)
sole trader (i)
’buy one get one free’ (BOGOF) (i)

Canadian Institute of Chartered Accountants (CICA) (i)


capital asset pricing model (CAPM) (i)
capital gearing ratio (i)
capital rationing (i)
capital turnover (i)
Carstensen, Laura (i)
cash budget (i), (ii)
cash conversion cycle (i), (ii)
lead time (i)
cash flow (i)
accounts payable, managing (i)
accounts receivable, managing (i)
credit setting and control (i), (ii)
debt collection (i)
debt factoring (i)
invoice discounting (i)
asset financing (i)
finance lease (i)
hire purchase (HP) (i)
operating lease (i)
Baumol–Tobin model (i), (ii), (iii)
worked example (i)
cash budget (i), (ii)
cash conversion cycle (i), (ii)
lead time (i)
cash-flow forecast, interpreting a (i)
compensating balances motive (i)
comprehension questions (i)
answers (i)
exercises (i)
answers (i)
financing cash flow (i)
importance of (i)
inventory management (i)
ABC inventory management (i), (ii)
economic order quantity (EOQ) model (i)
inventory turnover ratio (i)
just-in-time (JIT) (i)
reorder levels (i), (ii)
investment cash flow (i)
Miller–Orr model (i), (ii)
worked example (i)
operating cash flow (i)
opportunity cost (i), (ii)
overtrading (i), (ii)
precautionary motive (i)
speculative motive (i)
transaction motive (i)
worked examples (i), (ii), (iii), (iv), (v), (vi), (vii)
Chartered Accountants Ireland (CAI) (i)
Chartered Institute of Management Accountants (CIMA) (i), (ii)
Chartered Institute of Public Finance and Accountancy (CIPFA) (i)
Chen, Kung and Shimerda, Thomas (i)
Chiarello, Michael (i)
Circle (i)
Cisco Systems (i)
Closed Loop, The (i)
Coca-Cola (i)
CompareNet (i)
competitive advantage analysis (i)
competitive pricing (i)
concepts and systems (i)
accruals concept (i)
comprehension questions (i)
answers (i)
conservativism (i)
cost of sales working (i)
disaggregation (i)
double-entry bookkeeping (i)
exercises (i)
answers (i)
going concern (i)
materiality (i)
measurement rules (i)
neutrality (i)
property, plant and equipment
alternative depreciation methods (i)
non-current assets, disposal of (i)
recognition and subsequent measurement (i)
worked example (i), (ii), (iii)
prudence (i)
purpose of (i)
record keeping (i), (ii)
statement of cash flows (i), (ii)
estimates (i)
timing (i)
transactions (i)
working capital (i), (ii)
statement of comprehensive income (i), (ii)
gross profit (i)
operating profit (i)
profit for the year (i)
statement of financial position (balance sheet) (i)
current assets (i)
current liabilities (i)
non-current assets (i)
non-current liabilities (i)
worked examples (i), (ii), (iii), (iv)
Conceptual Framework for Financial Reporting (F) (i), (ii)
conservativism (i)
Consortium of Advanced Management, International (CAM–I) (i)
Consultative Committee of Accountancy Bodies (CCAB) (i)
consumer behaviour and perceptions (i)
ethics (i)
marketing mix, the (i)
perception of value, quality and difference (i)
product positioning (i)
controllability principle (i)
Co-operative Bank (i)
corporate social responsibility (CSR) reporting (i)
cost centres (i), (ii)
cost of sales working (i)
cost pools (i)
Costa (i)
cost-driver analysis (i)
cost-volume-profit analysis (i)
advertising (i)
break-even point (BEP) (i), (ii)
change in production systems (i)
change in selling price (i)
equation (i)
fixed costs (i), (ii)
margin of safety (i), (ii)
revenue (i), (ii)
total costs (i), (ii)
variable costs (i), (ii)
worked examples (i), (ii), (iii), (iv)
Council Directive 78/660/EEC (Fourth EU Directive on Company Law) (i)
Cruise, Tom (i)
current assets (i)
current cost accounting (i)
current liabilities (i)
current ratio (i)
Cutifani, Mark (i)

dangerous stakeholder (i)


debt collection (i)
debt factoring (i)
definitive stakeholder (i)
Dell (i)
Deloitte (i), (ii)
Delta Airlines (i)
demanding stakeholder (i)
dependent stakeholder (i)
depreciation (i)
disaggregation (i)
discounted cash flow (DCF) (i)
compounding (i)
discounting (i)
internal rate of return (IRR) (i)
net present value (NPV) (i), (ii)
discretionary stakeholder (i)
Disney (i)
dividend cover (i)
dividend yield (i)
dominant stakeholder (i)
dormant stakeholder (i)
double-entry bookkeeping (i)
Dun and Bradstreet (i)

earnings announcements (i)


earnings per share (EPS) (i)
economic order quantity (EOQ) model (i)
economic value added (EVA) (i), (ii)
EDF Energy (i)
efficiency ratios (i)
inventories holding period (i)
trade payables payment period (i)
trade receivables collection period (i)
Emirates Airlines (i)
Ernst and Young (EY) (i)
European Financial Reporting Advisory Group (EFRAG) (i)
Execution Premium, The (i)

Facebook (i)
finance lease (i)
financial accountants (i), (ii), (iii)
Big Four firms (i)
function on (i)
qualifications (i)
services provided by (i)
users of (i)
financial analysis (i)
comprehension questions (i)
answers (i)
exercises (i)
answers (i)
financial ratios, overview (i)
horizontal analysis (i), (ii)
ratio analysis
acid test ratio (i)
capital gearing ratio (i)
current ratio (i)
dividend cover (i)
dividend yield (i)
earnings per share (EPS) (i)
efficiency ratios (i)
gross profit margin (i)
interest cover (i)
introduction to (i), (ii)
inventories holding period (i)
investor ratios (i)
liquidity ratios (i)
net profit margin (i)
price to earnings ratio (P/E ratio) (i)
profitability ratios (i)
return on capital employed (ROCE) (i)
solvency ratios (i)
trade payables payment period (i)
trade receivables collection period (i)
weaknesses and limitations (i)
reasons for and user needs (i), (ii)
trend analysis (i), (ii)
vertical analysis (i), (ii), (iii)
financial budget (i)
Financial Services Authority (FSA) (i)
financial statements, overview (i)
fit together, how they (i), (ii)
problems, key (i)
required information (i)
terminology (i)
Tesco example (i), (ii), (iii), (iv), (v), (vi)
Firefly (i)
full-cost-plus pricing (i)

Gallagher, Mike and Paul, Beth (i)


going concern (i)
Google (i), (ii)
Gordon, Ian (i)
Graham’s The Family Dairy (i)
gross profit (i)
gross profit margin (i)
Guillon, Bruno (i)

HBoS (i)
hire purchase (HP) (i)
historical cost accounting (i)
Hotmail (i)

IASplus.com (i)
IBM (i)
In Search of Excellence (i)
InBev (i)
inflation (i)
comprehension questions (i)
answers (i)
corporate social responsibility (CSR) reporting (i)
earnings announcements (i)
information, demand for (i)
investor presentation (i)
press coverage (i)
social media and internet bulletins (i)
stakeholder management (i), (ii)
dangerous stakeholder (i)
definitive stakeholder (i)
demanding stakeholder (i)
dependent stakeholder (i)
discretionary stakeholder (i)
dominant stakeholder (i)
dormant stakeholder (i)
ING Direct (i)
Institute of Chartered Accountants in Australia (ICAA) (i)
Institute of Chartered Accountants in England and Wales (ICAEW) (i)
Institute of Chartered Accountants in New Zealand (ICANZ) (i)
Institute of Chartered Accountants in Scotland (ICAS) (i)
Institute of Chartered Accountants in Zimbabwe (ICAZ) (i)
interest cover (i)
internal rate of return (IRR) (i)
International Accounting Standards
IAS 1 Presentation of Financial Statements (i)
IAS 16 Property, Plant and Equipment (i), (ii)
IAS 33 Earning per Share (i)
IAS 7 Statement of Cash Flows (i)
International Accounting Standards Board (IASB) (i), (ii), (iii), (iv), (v), (vi), (vii), (viii), (ix)
International Federation of Accountants (IFAC) (i)
International Financial Reporting Standards (IFRSs) (i), (ii), (iii)
IFRS Foundation (i), (ii)
IFRS Interpretations Committee (i)
inventories holding period (i)
inventory management (i)
ABC inventory management (i), (ii)
economic order quantity (EOQ) model (i)
inventory turnover ratio (i)
just-in-time (JIT) (i)
reorder levels (i), (ii)
investment appraisal (i)
accounting rate of return (ARR) (i)
annuities (i)
capital asset pricing model (CAPM) (i)
capital rationing (i)
competitive advantage analysis (i)
comprehension questions (i)
answers (i)
cost-driver analysis (i)
depreciation (i)
discounted cash flow (DCF) (i)
compounding (i)
discounting (i)
internal rate of return (IRR) (i)
net present value (NPV) (i), (ii)
exercises (i)
answers (i)
inflation (i)
money method (i)
Monte Carlo simulation method (i)
opportunity costs (i)
payback period (PP) (i)
probability analysis (i)
qualitative considerations (i)
real method (i)
real options (i)
replacement of assets (i)
risk, return, and timing (i)
scenario analysis (i)
sensitivity analysis (i)
taxation (i)
value chain analysis (i)
weighted average cost of capital (WACC) (i)
worked examples (i), (ii), (iii), (iv), (v), (vi), (vii), (viii), (ix)
investment cash flow (i)
investment centre (i)
investor ratios (i)
dividend yield (i)
earnings per share (EPS) (i)
price to earnings ratio (P/E ratio) (i)
invoice discounting (i)

Johnson, H Thomas and Kaplan, Robert (i)


Jolie, Angelina (i)
just-in-time (JIT) (i), (ii), (iii)

Kaizen (i)
Kaplan, Robert (i)
Johnson, H Thomas and Kaplan, Robert (i)
Kaplan, Robert and Norton, David (i)
KPMG (i), (ii)

lean manufacturing (i)


Lehman Brothers (i)
Libby, Theresa and Lindsay, Murray (i)
life-cycle costing and pricing (i), (ii), (iii)
limited company (i), (ii)
liquidity ratios (i)
acid test ratio (i)
current ratio (i)
loss leaders (i)
Lucasfilm (i)

margin (i)
margin of safety (i), (ii)
marginal-cost-plus pricing (i)
market skimming pricing (i)
market types (i)
monopolistic competition (i)
monopoly (i)
oligopoly (i)
perfect competition (i)
marketing mix, the (i)
mark-up (i)
master budget (i)
materiality (i), (ii)
McCarthy, E Jerome (i)
merchandizing businesses (i)
MG Rover (i)
Microsoft (i)
Xbox (i) (ii)
Miller–Orr model (i), (ii)
worked example (i)
Mitchell, Ronald, Agle, Bradley and Wood, Donna (i)
mixed measurement model (i)
money measurement concept (i)
money method (i)
monopolistic competition (i)
monopoly (i)
Monte Carlo simulation method (i)
Moody’s (i)
Morrison, Diane (i)
Morrisons (i)
Mulberry (i)
MV Agusta (i)
net present value (NPV) (i), (ii)
net profit margin (i)
neutrality (i)
Next (i)
Nintendo (i)
Nokia (i)
non-current assets (i)
non-current liabilities (i)
Northern Rock (i)
not-for-profits (i), (ii), (iii)
3Es (i)
example balance scorecard (i)

oligopoly (i)
one-off pricing (i)
operating cash flow (i)
operating lease (i)
operating profit (i)
operational budget (i)
operational decision making (i)
comprehension questions (i)
answers (i)
cost-volume-profit analysis (i)
advertising (i)
break-even point (BEP) (i), (ii)
change in production systems (i)
change in selling price (i)
equation (i)
fixed costs (i), (ii)
margin of safety (i), (ii)
revenue (i), (ii)
total costs (i), (ii)
variable costs (i), (ii)
worked examples (i), (ii), (iii), (iv)
exercises (i)
answers (i)
relevant costing (i), (ii)
opportunity costs (i)
outsourcing (i)
qualitative factors (i)
shutting down business segments (i)
sunk cost (i)
worked examples (i), (ii),(iii)
steps in (i), (ii)
operational performance management (i)
opportunity cost (i), (ii), (iii), (iv)
optional pricing (i)
outsourcing (i)
overhead absorption rate (OAR) (i)
overtrading (i), (ii)
Oxford Instruments plc (i)

Pacioli, Luca (i)


Page, Larry (i)
Pareto Principle (i)
Parsa, Ali (i)
partnership (i)
payback period (PP) (i)
penetration pricing (i)
perfect competition (i)
performance management (i)
balanced scorecard (i), (ii)
behavioural risks of (i)
comprehension questions (i)
answers (i)
controllability principle (i)
exercises (i)
answers (i)
just-in-time (JIT) (i), (ii)
life-cycle costing (i)
3Es (i)
not-for-profits (i), (ii)
example balance scorecard (i)
operational performance management (i)
performance indicators
capital turnover (i)
economic value added (EVA) (i), (ii)
non-financial (i)
residual income (RI) (i)
return on investment (ROI) (i), (ii)
performance reporting (i)
PEST analysis (i), (ii)
public sector businesses (i)
responsibility centres
cost centres (i)
investment centre (i)
profit centre (i)
revenue centre (i)
standard costing (i)
strategic performance management (i)
target costing (i)
throughput accounting (i)
total quality management (TQM) (i), (ii)
variance analysis (i), (ii)
fixed overhead variance (i)
labour variance (i)
materials variance (i)
rate variance (i)
sales variance (i)
variable overhead variance (i)
volume variance (i)
worked examples (i), (ii), (iii)
PEST analysis (i), (ii)
Peters, Tom (i)
Pitt, Brad (i)
Porter, Michael (i)
predatory pricing (i)
premium pricing (i)
price elasticity of demand (PED) (i), (ii)
factors determining (i)
price to earnings ratio (P/E ratio) (i)
PricewaterhouseCoopers (i), (ii)
pricing (i)
3Cs (i), (ii)
absorption costing (i)
activity-based costing (ABC) and pricing (i)
bundle pricing (i)
competitive pricing (i)
comprehension questions (i)
answers (i)
consumer behaviour and perceptions (i)
ethics (i)
marketing mix, the (i)
perception of value, quality and difference (i)
product positioning (i)
cost centre (i)
cost pools (i)
exercises (i)
answers (i)
full-cost-plus pricing (i)
Kaizen (i)
life-cycle costing and pricing (i), (ii)
marginal-cost-plus pricing (i)
market skimming pricing (i)
market types (i)
monopolistic competition (i)
monopoly (i)
oligopoly (i)
perfect competition (i)
mark-up vs margin (i)
new vs existing market (i)
new vs existing product (i)
one-off pricing (i)
optional pricing (i)
overhead absorption rate (OAR) (i)
penetration pricing (i)
premium pricing (i)
price elasticity of demand (PED) (i), (ii)
factors determining (i)
stratified pricing (i)
target pricing and costing (i), (ii)
value engineering (i)
value pricing (i)
volume- vs price-driven strategy (i)
worked examples (i), (ii), (iii), (iv)
probability analysis (i)
product positioning (i)
production businesses (i)
profit centre (i)
profitability ratios (i)
gross profit margin (i)
net profit margin (i)
return on capital employed (ROCE) (i)
property, plant and equipment
alternative depreciation methods (i)
non-current assets, disposal of (i)
recognition and subsequent measurement (i)
worked example (i), (ii), (iii)
prudence (i)
psychological pricing (i)
public sector businesses (i)
Pyhrr, Peter (i)
ratio analysis
efficiency ratios (i)
inventories holding period (i)
trade payables payment period (i)
trade receivables collection period (i)
introduction to (i), (ii)
investor ratios (i)
dividend yield (i)
earnings per share (EPS) (i)
price to earnings ratio (P/E ratio) (i)
liquidity ratios (i)
acid test ratio (i)
current ratio (i)
profitability ratios (i)
gross profit margin (i)
net profit margin (i)
return on capital employed (ROCE) (i)
solvency ratios (i)
capital gearing ratio (i)
dividend cover (i)
interest cover (i)
weaknesses and limitations (i)
real method (i)
real options (i)
regulation (i), (ii), (iii)
Conceptual Framework (F) (i), (ii)
IFRS Foundation (i), (ii)
IFRS Interpretations Committee (i)
International Accounting Standards Board (IASB) (i), (ii)
principles (i)
Relevance Lost (i)
relevant costing (i), (ii)
opportunity costs (i)
outsourcing (i)
qualitative factors (i)
shutting down business segments (i)
sunk cost (i)
worked examples (i), (ii),(iii)
residual income (RI) (i)
responsibility centres
cost centres (i)
investment centre (i)
profit centre (i)
revenue centre (i)
return on capital employed (ROCE) (i)
return on investment (ROI) (i), (ii)
revenue centre (i)
Rio Tinto (i)
rolling budgets (i)
Ryanair (i), (ii)
Chairman’s report (i)
efficiency (i)
inventories holding period (i)
trade payables payment period (i)
trade receivables collection period (i)
horizontal analysis and trend analysis (i), (ii)
investment (i)
dividend yield (i)
earnings per share (EPS) (i)
price to earnings ratio (P/E ratio) (i)
liquidity (i)
acid test ratio (i)
current ratio (i)
profitability (i)
gross profit margin (i)
net profit margin (i)
solvency (i)
capital gearing ratio (i)
dividend cover (i)
interest cover (i)
vertical analysis (i), (ii), (iii)

Sainsbury (i)
Samsung (i), (ii)
scenario analysis (i)
Schulz, Don, Tannenbaum, Stanley and Lauterbon, Robert (i)
Schumacher, Michael (i)
sensitivity analysis (i)
service businesses (i)
Shimizu, Koichi (i)
Skype (i)
sole trader (i)
solvency ratios (i)
capital gearing ratio (i)
dividend cover (i)
interest cover (i)
Sony (i)
South African Institute of Chartered Accountants (SAICA) (i)
Spanair (i)
stakeholder management (i), (ii)
dangerous stakeholder (i)
definitive stakeholder (i)
demanding stakeholder (i)
dependent stakeholder (i)
discretionary stakeholder (i)
dominant stakeholder (i)
dormant stakeholder (i)
standard costing (i)
Standard Life (i)
Starbucks (i)
statement of cash flows (i), (ii)
estimates (i)
timing (i)
transactions (i)
working capital (i), (ii)
statement of comprehensive income (i), (ii)
gross profit (i)
operating profit (i)
profit for the year (i)
statement of financial position (balance sheet) (i)
current assets (i)
current liabilities (i)
non-current assets (i)
non-current liabilities (i)
Statoil (i)
Stern Steward & Co (i)
strategic performance management (i)
stratified pricing (i)
Summa de arithmetica, geometria, proportioni et proportionalita (i)
sunk cost (i)
Sutherland, Euan (i)

TagHeuer (i)
target pricing and costing (i), (ii), (iii)
taxation (i)
Tesco plc (i), (ii), (iii), (iv)
Texas Instruments (i)
throughput accounting (i)
Tootell, Barry (i)
total quality management (TQM) (i), (ii)
Toyota (i), (ii)
trade payables payment period (i)
trade receivables collection period (i)
trend analysis (i), (ii)
Twitter (i)

UK Competition Commission (UKCC) (i)


Underwood, Toby (i)

value chain analysis (i)


value engineering (i)
value pricing (i)
variance analysis (i), (ii)
fixed overhead variance (i)
labour variance (i)
materials variance (i)
rate variance (i)
sales variance (i)
variable overhead variance (i)
volume variance (i)
vertical analysis (i), (ii), (iii)
Vodafone (i)

Waterman, Robert (i)


weighted average cost of capital (WACC) (i)
Wet ‘N’ Wild (i)
Whitbread plc (i)
working capital (i), (ii)

Yammer (i)
YouTube (i)
zero-based budgets (i)
Publisher’s note
Every possible effort has been made to ensure that the information contained in this book is accurate at the time
of going to press, and the publishers and authors cannot accept responsibility for any errors or omissions,
however caused. No responsibility for loss or damage occasioned to any person acting, or refraining from action,
as a result of the material in this publication can be accepted by the editor, the publisher or either of the authors.

First published in Great Britain and the United States in 2014 by Kogan Page Limited

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the
Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form
or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in
accordance with the terms and licences issued by the CLA. Enquiries concerning reproduction outside these terms
should be sent to the publishers at the undermentioned addresses:

2nd Floor, 45 Gee Street 4737/23 Ansari Road


1518 Walnut Street, Suite 1100
London EC1V 3RS Daryaganj
Philadelphia PA 19102
United Kingdom New Delhi 110002
USA
www.koganpage.com India

© Matt Bamber and Simon Parry, 2014

The right of Matt Bamber and Simon Parry to be identified as the author of this work has been asserted by them in
accordance with the Copyright, Designs and Patents Act 1988.

ISBN 978 0 7494 6913 9


E-ISBN 978 0 7494 6914 6

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library.

CIP data is available.


Library of Congress Control Number: 2014004436

Typeset and eBook by Graphicraft Limited, Hong Kong


Print production managed by Jellyfish
Printed and bound by CPI Group (UK) Ltd, Croydon CR0 4YY

You might also like