Exclusive use

Subject CP1
Revision Notes
For the 2019 exams

Modelling
Booklet 5

Covering

Chapter 17 Modelling
Chapter 18 Data
Chapter 19 Setting assumptions

The Actuarial Education Company

Exclusive use
 Exclusive use

CONTENTS

Contents Page
Links to the Course Notes and Syllabus 2
Overview 4
Brainstorm 7
Core Reading Questions 8
Solutions to Core Reading Questions 15
Past Exam Questions 50
Solutions to Past Exam Questions 70
Brainstorm (2) 147
Brainstorm – solution 148
Final comments 149
Exam Preparation Checklist 158

Copyright agreement

All of this material is copyright. The copyright belongs to Institute and

Faculty Education Ltd, a subsidiary of the Institute and Faculty of Actuaries.
The material is sold to you for your own exclusive use. You may not hire
out, lend, give, sell, transmit electronically, store electronically or photocopy
any part of it. You must take care of your material to ensure it is not used or
copied by anyone at any time.

Legal action will be taken if these terms are infringed. In addition, we may
seek to take disciplinary action through the profession or through your
employer.

These conditions remain in force after you have finished using the course.

© IFE: 2019 Examinations Page 1

 Exclusive use

LINKS TO THE COURSE NOTES AND SYLLABUS

Material covered in this booklet

Chapter 17 Modelling
Chapter 18 Data
Chapter 19 Setting assumptions

These chapter numbers refer to the 2019 edition of the ActEd Course Notes.

Syllabus objectives covered in this booklet

10 Specifying the problem

10.2 Data (Chapter 18)

10.2.1 Explain the ethical and regulatory issues involved in working

with personal data and extremely large data sets.

10.2.2 Explain the main issues to be addressed by a data governance

policy and its importance for an organisation.

10.2.3 Explain the risks associated with use of data (including

algorithmic decision making).

10.2.4 Discuss the data requirements for determining values for

assets, future benefits and future funding requirements.

10.2.5 Describe the checks that can and should be made on data.

10.2.6 Describe the circumstances under which the ideal data

required might not be available and discuss ways in which this
problem may be overcome.

10.2.7 Describe how to determine the appropriate grouping of data to

achieve the optimal level of homogeneity.

11 Producing the solution

11.1 Modelling (Chapter 17)

11.1.1 Describe the approaches available to produce the solution to

an actuarial or financial problem.

Page 2 © IFE: 2019 Examinations

 Exclusive use

11.1.2 Describe the construction of actuarial models to produce

solutions in terms of:
 the objectives of the model
 the operational issues that should be considered in
designing and running models.

11.1.3 Describe the use of models for:

 pricing or setting future financing strategies
 risk management: assessing the capital requirements and
the return on capital or the funding levels required
 assessing the provisions needed for existing commitments
to provide benefits on contingent events
 pricing and valuing options and guarantees.

11.1.4 Describe how sensitivity analysis of the results of the models

can be used to help decision making.

11.2 Assumption setting (Chapter 19)

Describe the principles behind the determination of

assumptions as input to a model relevant to producing a
specific solution having regard to:
 the types of information that may be available to help in
determining the assumptions to be used
 the extent to which each type of information may be useful,
and the other considerations that may be taken into
account, in deciding the assumptions
 the level of prudence in the assumptions required to meet
the objectives of the client.

© IFE: 2019 Examinations Page 3

 Exclusive use

OVERVIEW

The material covered in this booklet addresses Syllabus objectives relating

to the Producing the solution stage of the Actuarial Control Cycle, and part of
Specifying the problem stage.

In this booklet, we look at a typical solution to an actuarial problem – building

a model. This model may be deterministic or stochastic in nature, and the
Core Reading looks at the pros and cons of each.

This booklet also considers the various inputs into that model:
 policy data, and the checks that can be performed to ensure that this
data is accurate
 assumptions, eg claims, expenses, investment returns, withdrawal
rates.

Modelling

A model is one of the tools that actuaries have available to ensure that the
advice they provide is as good as possible. The various uses to which
models can be put, in the worlds of insurance and pensions, include:
 contract design
 setting premiums or contributions
 assessing the return on capital
 assessing surplus and analysing its components
 assessing capital requirements
 assessing the financial condition of an insurance company and future
prospects
 costing options and guarantees
 determining discontinuance terms
 assessing the impact of decisions, eg an investment strategy, a
reinsurance program.

As well as the technical areas of model design, there are operational issues
that need to be considered, such as ease of communication of results, ease
of validation, documentation, and frequency of cashflow projection.

Page 4 © IFE: 2019 Examinations

 Exclusive use

A model should be ‘dynamic’, ie allow appropriately for the interaction of

variables. For example, investment returns should be consistent with
inflation, and salary growth should be consistent with economic growth.

A decision needs to be made as to whether to use a deterministic or

stochastic model. Stochastic models are typically used to assess the impact
of financial guarantees / options and when we are interested in events
affecting the tails of distributions, eg setting a reinsurance retention level,
asset-liability modelling. Typically, deterministic models with scenario testing
have been used for pricing, profitability and capital work but, as computer
power increases, many providers are now also using stochastic models.

Policy data

Financial providers need to use personal data relating to their customers or

members, so it is important to understand the ethical and legislative issues
that arise when working with such data.

A model will require accurate and representative model points representing

policy or member data. The actuary can carry out a range of checks to be
satisfied as to the quality of the policy data. These checks fall into the
following main categories:
 reconciliation of movements
 consistency (with previous data) and reasonableness checks
 spot checks and checks on unusual values
 cross checks with other sources of data.

We also look at the various sources of data when the company has no data
relating to the contract in question. Such sources include looking at similar
contracts, industry data, reinsurers, actuarial tables, national and overseas
statistics. The key risks that face an actuary when using data are also
covered.

Assumptions

When setting assumptions it is important to consider the use to which they

will be put, eg for pricing or provisioning, and the needs of the client.
Particular care must be taken over the choice of those assumptions which
will have the most financial significance. It is also important to allow for any
consistency that should exist between the various assumptions and any
regulation prescribing or influencing the assumptions to be used.

© IFE: 2019 Examinations Page 5

 Exclusive use

Values need to be assigned to assumptions such as investment return,

expenses, withdrawal rates and claim rates. These values will reflect (but
not necessarily equal) the expected future experience of the parameters.
The estimates for demographic assumptions will usually be based on
relevant historical data, but adjusted for trends and any known changes in
circumstances.

Page 6 © IFE: 2019 Examinations

 Exclusive use

BRAINSTORM

How much can you remember from the Course Notes about modelling? Use
this page to brainstorm the topic, writing down as many key words as you
can think of. You should spend at least five minutes thinking about the
major ideas before checking your answer with the solution at the back of this
booklet (which includes over 40 ideas).

Requirements

Modelling

Uses
Steps

Cashflows

© IFE: 2019 Examinations Page 7

 Exclusive use

CORE READING QUESTIONS

The following questions will test your knowledge of the Core Reading
covered by this booklet. We have also included the relevant terms from the
Glossary. By attempting every question and reading every solution carefully,
you will cover the full Core Reading for this part of the course.

The number of marks is given to indicate the amount of detail required in the
answer.

Chapter 17 – Modelling

1 Describe why the solution to many actuarial problems involves building

a model. [3]

2 What is a model? [2]

3 List three potential approaches to obtaining a model and list five criteria
that should be considered in deciding between these approaches. [4]

4 State a type of model of which there are now many in existence. For
what variables are there fewer models available? [2]

5 Outline the key objective when constructing an actuarial model,

including when this objective is particularly relevant. [2]

6 Describe nine operational issues that need to be considered in relation

to model design and construction. [6]

7 Discuss the merits of deterministic and stochastic models. [4]

8 Explain, with an example relating to pricing an investment guarantee,

how a combination of a stochastic and deterministic model could be
used. [2]

9 Outline with examples how a model can be made ‘dynamic’. [3]

10 Give an example of a key use of a model in an insurance company. [1]

11 Explain the use of ‘model points’ in a model and how suitable model
points can be obtained for use in product pricing. [4]

12 Outline what ‘running a pricing model’ with model points actually entails.
[1]

Page 8 © IFE: 2019 Examinations

 Exclusive use

13 Outline how the risk discount rate for the model can be determined. [4]

14 In a model used to set premiums or charges, how will the key objective
of the model be expressed? [1]

15 State five things that might need to be reconsidered in the process of

pricing a product, as a result of the premiums or charges produced have
been considered for marketability. [2½]

16 Discuss how and why a model of the business of the whole company
will be used in the process of pricing a new product. [3]

17 State what has to be done once appropriate prices have been

determined for the model points that were chosen for the pricing
process. [1]

18 Outline how the capital requirement and return on capital would be

assessed for a new product. [2]

19 For a benefit scheme, the equivalent to determining the price for a

product is to set the future financing strategy. Describe how a model
would be used to set a future financing strategy. [4]

20 Describe the use of models for risk management. [3]

21 Explain why the use of models is limited in relation to valuing the

liabilities of a life insurance company or pension scheme. [2]

22 Outline how a stochastic model can be used when determining

provisions for existing commitments. [1]

23 Outline the key modelling issues in relation to valuing options and

guarantees. [3]

24 Describe how the variability of modelling results can be illustrated. [4]

25 Outline two types of errors that can arise in a model and how they can
be investigated. [2]

26 Describe how the variability of modelling results can be used to help in:
 pricing [1]
 assessing the return on capital or profitability of business. [1]

27 Outline four ways of allowing for statistical risk in a model. [2]

© IFE: 2019 Examinations Page 9

 Exclusive use

28 A unit-linked life assurance policy guarantees to pay a maturity value

equal to the sum of premiums on the chosen maturity date, or the value
of units allocated if greater. At all other times the surrender value is
based on the value of units.

Describe the steps involved in assessing the provision to be made for

the cost of this guarantee. [10]

Chapter 18 – Data

29 Explain why the use of data is an increasingly significant issue for

organisations. [2]

30 Define personal data. [1]

31 What is the main ethical responsibility for organisations in relation to

personal data? [1]

32 Explain the purpose of data protection legislation, giving examples of

how it compares between different countries. [4]

33 State the eight principles of the UK’s Data Protection Act that relate to
processing personal data. [4]

34 Give three examples of possible consequences of non-compliance with

data protection legislation when processing personal data. [2]

35 Explain the relevance of anonymity to the definition of personal data. [2]

36 Give two examples of competition legislation that might limit the uses to
which data can be put. [2]

37 Outline the potential consequences of non-compliance with this

legislation. [1]

38 List seven examples of information that can constitute sensitive

personal data. [3½]

39 State whether sensitive personal data is generally subject to more or

less strict regulation than ordinary personal data. [1]

40 Give four examples of conditions that might have to be satisfied in order

to permit sensitive personal data to be processed. [2]

41 Explain what is meant by ‘big data’, including its key characteristics. [2]

Page 10 © IFE: 2019 Examinations

 Exclusive use

42 Describe the main data protection considerations for organisations

using big data. [8]

43 Define data governance. [1]

44 Describe the purpose and typical contents of a data governance policy.

[4]

45 State four risks to which organisations are exposed if they do not have
adequate data governance procedures. [2]

46 Describe the key data issues when businesses are combined by merger
or takeover. [4]

47 Give four examples of possible risks that arise when using data, which
relate to data volume or quality. [4]

48 List eight reasons why historical data may not be a good reflection of
future experience. [4]

49 Outline the two risks that arise where an actuary attempts to group data
into broadly homogeneous groups. [1]

50 Give three examples of other risks that are associated with the use of
data. [2]

51 Explain what is meant in general by electronic or automated trading,

including its advantages. [2]

52 Explain what is specifically meant by algorithmic trading. [2]

53 Describe four risks that are associated with algorithmic trading. [4]

54 Explain what is meant by programmed trading. [1]

55 Programmed trading example

You own 100 shares of a company, with each share currently equal to
£1. The rules in place for buying/selling the stock require you to sell the
stock when it falls in value by 5% from the starting point and buy when
the stock rises in value by 5% above its low point.

Say the share price falls from £1.00 to £0.93 on a given day and then
bounces back to £1.05 before the end of the day.

What are your cashflows? [3]

© IFE: 2019 Examinations Page 11

 Exclusive use

56 Discuss the pros and cons of the programmed trading strategy in the
previous example. [5]

57 When data is required for a number of tasks, what is a key principle in

its provision? [1]

58 Comment on how different sources of data (ie publicly available and

internal) may be suitable for different purposes. [2]

59 Outline two sources of problems of data quality and quantity and why
the current management are not necessarily to blame for these
problems. [2]

60 Describe how it can be ensured that good quality data is obtained from
proposal and claim forms. [4]

61 Discuss the data issues, data requirements and data sources in benefit
scheme valuations. [7]

62 List five assertions regarding data that an actuary should aim to check.
[2]

63 Outline eleven data checks that might be carried out on valuation data.
[5½]

64 Give examples of two sets of circumstances where data will not be

‘ideal’. [4]

65 Describe the drawbacks of using summarised data for valuation

purposes. [2]

66 Describe, with examples, what is meant by ‘industry-wide’ data

collection schemes. [3]

67 Discuss the advantages of having access to industry-wide data. [3]

68 Discuss the disadvantages of having access to industry-wide data. [5]

69 List two further examples of sources of data for an insurance company.

[1]

70 Describe the process of risk classification. [5]

Page 12 © IFE: 2019 Examinations

 Exclusive use

Chapter 19 – Setting assumptions

71 Outline five key considerations in setting assumptions for actuarial work.

[2½]

72 Give four examples of where historical data can be useful in setting

future assumptions. [4]

73 Give three examples of where current data can be useful in setting

future assumptions. [2]

74 In some instances, what will assumptions be defined by? [1]

75 Explain why is it usually inappropriate to take an average rate from the

past data and use this for projecting into the future. [3]

76 When analysing past data, what must the relevance of that past data to
future projections be balanced against? [1]

77 List seven possible features of past data that the actuary may need to
adjust for. [3]

78 Why is past economic data usually unsuitable for estimating future

levels? Why is it not sensible to use economic data from a short
time-frame? How might past data be amended to overcome these
problems? [2]

79 What might past data for price inflation be used as an indicator for?
What measure will give a better indicator of future price inflation? [3]

80 Give an example of when past price inflation data can be useful for
setting assumptions. [1]

81 Give an example of a fiscal change that should be allowed for when

using historical dividend levels to set assumptions for future dividend
levels. Explain a potential problem that arises if an explicit adjustment is
made. [1]

82 What significant factor should be borne in mind when projecting

mortality experience into the future? How does this affect the analysis?
[2]

83 Why is it important to split data into homogeneous groups? In practice,

how is historical data used? [4]

© IFE: 2019 Examinations Page 13

 Exclusive use

84 What particular care needs to be taken when accessing older data

produced either by the State or by companies? [3]

85 Give two examples of one-off impacts that might invalidate the use of
past data. [1]

86 Describe standard tables based on population data and one of their

disadvantages. [2]

87 Outline the work of the Continuous Mortality Investigation Bureau in the

UK. [1]

88 What considerations should be taken into account when using standard

tables? [1]

89 Describe three factors to consider when deciding on the need for

accuracy in setting assumptions. [3]

90 Discuss the assumptions to use when acting as expert witness to

determine a fair compensation settlement between two parties. [2]

91 Give two examples of implicit assumptions within a model, in relation to

the funding of an occupational pension scheme. [2]

92 State three ways of allowing for risk in the assumptions. [1½]

93 In pricing a contract, what is another assumption that will have to be

made in order to determine whether the contract is a good investment?
[1]

94 What factors make a contract design riskier? [3]

95 Describe the term ‘profit criterion’. [3]

Chapter 39 – Glossary

96 Define the following terms:

 Best estimate
 Internal rate of return
 Profit test
 Risk discount rate.

Page 14 © IFE: 2019 Examinations

 Exclusive use

SOLUTIONS TO CORE READING QUESTIONS

This section contains the answers to the Core Reading Questions.

The text given in Arial Bold font is Core Reading.

The text given in Arial Bold Italic font is additional Core Reading that is not
directly related to the answers to specific Core Reading Questions.
____________

Chapter 17 – Modelling

1 There are various approaches that can be taken to produce the

solution to an actuarial or financial problem. Simple problems can
have a simple solution that is arrived at by some straightforward
mathematics, for example calculating the yield on a fixed-interest
asset, or the present value of a series of known cashflows.

However, most problems that require actuarial skills involve taking a

view on uncertain future events. It is possible to take a view on various
parameters, such as future economic conditions, future mortality rates,
or the amount of business that a provider might write in future, and
produce a single answer that is appropriate in these best estimate
conditions. If this is done then the communication of the solution to
the client needs care, because of the uncertainties in the underlying
assumptions.

In these circumstances the client is likely to wish to know the

variability of the answer provided, should circumstances not be as
estimated. To assess the effects of varying the assumptions used in
producing the answer, it is normally necessary to use an actuarial
model of future events.
____________

2 A model can be defined as ‘a cut-down, simplified version of reality

that captures the essential features of a problem and aids
understanding’. The final phrase in this definition recognises the
importance of being able to communicate the results effectively.
Modelling requires a balance to be struck between realism (and hence
complexity) and simplicity (for ease of application, verification and
interpretation of results).
____________

© IFE: 2019 Examinations Page 15

 Exclusive use

3 When faced with an actuarial or financial problem, there are various

approaches to modelling:
 a commercial modelling product could be purchased
 an existing model could be reused, possibly after modification
 a new model could be developed.

The merits of each of these approaches will depend on the following:

 the level of accuracy required
 the ‘in-house’ expertise available
 the number of times the model is to be used
 the desired flexibility of the model
 the cost of each option.
____________

4 There are now many stochastic asset models in existence, in both the
public and private domains. There are fewer models available for other
variables, such as mortality and voluntary discontinuance, but these
are starting to be developed.
____________

5 Any model should be fit for the purpose for which it is being used.
This is particularly relevant when a model is being purchased from an
external provider or when an existing model is being reused for a
different purpose, possibly after modification. Even with new purpose-
built models, the potential for model error remains – a model that
replicates past results may still prove unreliable in projecting future
results.
____________

6 Operational issues
 The model being used should be adequately documented.
 The workings of the model should be easy to appreciate and
communicate. The results should be displayed clearly.
 The model should exhibit sensible joint behaviour of model
variables.
 The outputs from the model should be capable of independent
verification for reasonableness and should be communicable to
those to whom advice will be given.

Page 16 © IFE: 2019 Examinations

 Exclusive use

 The model, however, must not be overly complex so that either the
results become difficult to interpret and communicate or the model
becomes too long or expensive to run, unless this is required by
the purpose of the model. It is important to avoid the impression
that everything can be modelled.
 The model should be capable of development and refinement
– nothing complex can be successfully designed and built in a
single attempt.
 A range of methods of implementation should be available to
facilitate testing, parameterisation and focus of results.
 The more frequently the cashflows are calculated the more reliable
the output from the model, although there is a danger of spurious
accuracy.
 The less frequently the cashflows are calculated the faster the
model can be run and results obtained.
____________

It will be necessary to decide between deterministic and stochastic

modelling processes.
____________

7 A deterministic model is more readily explicable to a non-technical

audience, since the concept of variables as probability distributions is
not easy to understand. It is clearer what economic scenarios have
been tested. The model is usually cheaper and easier to design and
quicker to run. The disadvantages are that it requires thought as to the
range of economic scenarios that should be tested. Also, users can
get ‘blinded by science’ by complex models, assuming they must be
working correctly, but without verifying or testing this.

A stochastic model tests a wider range of economic scenarios. The

programming is more complex and the run time longer, but the benefit
is in the quality of the result. It does depend on the parameters that are
used in any standard investment model. Stochastic models are
particularly important in assessing the impact of financial guarantees
or to allow for investment mismatching risks.
____________

8 In many cases the problem can be solved by a combination of

stochastic and deterministic modelling. Variables whose performance
is unknown and where the risk associated with them is high might be
modelled stochastically, while other variables can sensibly be
modelled deterministically.

© IFE: 2019 Examinations Page 17

 Exclusive use

For example, a model for pricing an investment guarantee attached to a

life insurance policy might use a stochastic investment model, but
would be unlikely to model fluctuations in mortality rates other than
deterministically. This is because it is normally self-evident which
direction of movement in mortality rates would give rise to financial
difficulties.
____________

9 In all cases the dynamism of the model is vital. Rules need to be

determined as to how the various features would interact in different
circumstances. For example, how life assurance bonus rates would
vary with fixed-interest yields, how policy lapse rates would vary with
economic conditions, or how unemployment rates would vary with
economic conditions. These interactions are usually much more
important than the type of model.

Considerable actuarial judgement may be required in choosing and

using the model and in setting the parameters and interactions
between the different features.
____________

As well as the technical areas of model design, which are covered in

the Core Practices subjects, there are several operational issues that
need to be considered.
____________

10 A model could be developed to determine a premium or charging

structure for a new or existing product that will meet an insurance
company’s profit requirement.
____________

11 The underlying business being modelled will typically comprise a very

wide range of different policies, and these will need to be brought
together into a manageable number of relatively homogeneous groups.

The groupings need to be made in a way that each policy in a group is

expected to produce similar results when the model is run. It is then
sufficient for a representative single policy in each group to be run
through the model, the result to be found, and for this result to be
scaled up to give the result of the total set of policies in the group.

The representative single policy in a group is termed a ‘model point’

and a set of such ‘model points’ can then be used to represent the
whole of the underlying business.

Page 18 © IFE: 2019 Examinations

 Exclusive use

A set of model points will be chosen to represent the expected new

business under the product. In the case of an existing product, the
profile of the existing business, modified to allow for any expected
changes in future, can be used to obtain the model points. For a new
product, the profile of any similar existing product combined with
advice from the company’s marketing department would be used.
____________

12 For each model point, cashflows would be projected, allowing for

reserving and solvency margin requirements, on the basis of a set of
base values for the parameters in the model. The net projected
cashflows will then be discounted at a rate of interest, the risk discount
rate.
____________

13 This could be a rate that allows for:

 the return required by the company; and
 the level of statistical risk attaching to the cashflows under the
particular contract, ie their variation about the mean as
represented by the cashflows themselves.

The level of statistical risk could be assessed:

 in some situations, analytically – by considering the variances of
the individual parameter values used
 by using sensitivity analysis, as described below, with
deterministically assessed variations in the parameter values
 by using stochastic models for some or all of the parameter values
and simulation
 by comparison with any available market data.

Alternatively a stochastic discount rate could be used.

In theory, a separate risk discount rate should be applied to each

separate component of the cashflows, as the statistical risk associated
with each component will be different. In practice a single risk
discount rate is commonly used, bearing in mind the ‘average’ risk of
the product.
____________

14 The premium or charges for the model point can then be set so as to
produce the profit required by the company.
____________

© IFE: 2019 Examinations Page 19

 Exclusive use

15 The premiums, or charges, produced need to be considered for

marketability. This might lead to a reconsideration of:
 the design of the product, so as either to remove features that
increase the risk within the net cashflows, or to include features
that will differentiate the product from those of competing
companies
 the distribution channel to be used, if that would permit either a
revision of the assumptions to be used in the model, or a higher
premium or charges to be used without loss of marketability
 the company’s profit requirement
 the size of the market
 whether to proceed with marketing the product.
____________

16 The net cashflows in respect of the model points, appropriately scaled

up for the expected new business under the product, will be
incorporated into a model of the business of the whole company. It is
possible for the desired level of profitability to be reached in aggregate,
without requiring every single model point to be profitable in its own
right. If certain model points are unprofitable, the aggregate
profitability of the business is then exposed to changes in mix and
volume of the contracts sold.

The actuary can assess the impact on capital management of writing

the product, by observing the modelled amount and timing of
cashflows. If capital is a problem, this may lead to a reconsideration of
the design of the product to reduce or amend the timing of its financing
requirement.
____________

17 Once acceptable premiums or charges have been determined for the

model points, premiums or charges for all contract variations can be
determined.
____________

18 For a benefit scheme, the equivalent to determining the price for a

product is setting the future financing strategy, and similar modelling
techniques can be used.

The existing membership can be divided into categories and

represented by a set of model points. Similar potential new members
can be represented, perhaps by a single model point at the average
entry age and salary.

Page 20 © IFE: 2019 Examinations

 Exclusive use

A potential financing strategy is determined, in terms of both the

amount and timing of future contributions. The cashflows from the
existing assets and future contributions can be modelled, as can the
liability cashflows, taking all the possible decrements into account.

Unlike an insurance company, a benefit scheme can show a deficit at a

point in time (ie the value of accumulated assets does not exceed the
value of accrued liabilities), provided that there is a sponsor with a
good enough covenant to make good the shortfall. However, the
scheme does need to be solvent to the extent that it has sufficient
assets to meet benefit outgo as it falls due. A well-designed model will
check this feature as well as determining the discounted value of asset
and liability cashflows.

Considerations such as the choice of risk discount rate, and the need
to test sensitivities to changes in conditions are all similar to those in
product pricing.
____________

19 Cashflow models are used in risk management to determine the

amount of capital that it is necessary to hold to support the risks
retained by a financial institution.

(The various modelling approaches are discussed in a later booklet on

the risk management process.)

As well as the full corporate model to assess capital requirements,

models of specific risks can be used to determine the extent of a risk
event that will occur at a given probability, even if a full stochastic
model is too slow, too complex, or otherwise not used.

For example, a company that is targeting being able to withstand a

0.1% probability of ruin needs to know what equity market fall to test in
a deterministic scenario.

A standard equity market stochastic model can be used and calibrated

to historical performance of the market being considered. By running
the model several thousand times and ranking the results, the equity
fall that gives the one in a thousand worst result can be found.
____________

20 The net cashflows for the model points described in the section on
pricing above can be grossed up for the expected new business and
used to assess the amount of capital that will be required to write the
product either on a regulatory or an economic basis.

© IFE: 2019 Examinations Page 21

 Exclusive use

Any one-off development costs can be added, to the extent that they
are not amortised and included in the cashflows used. This gives the
total capital requirement and can be compared with the profits
expected to emerge from the product to determine the expected return
on that capital.
____________

21 The normal procedure for determining life assurance or pension

scheme liabilities is to value the benefits for each policy or scheme
member individually. In many territories this may be required by
legislation or regulation.

Consequently for published results there is little scope for using model
points. However, before finalising a published basis, many ‘what if’
questions might be asked. These could be answered by running a
model of the business. For smaller schemes or sections of a
company’s business it might be just as quick to run the whole data file
to answer the question and eliminate the model risk, given the current
speed of computers.
____________

22 As part of assessing a realistic provision it is necessary to consider

the effect of changes in economic scenarios. For example, using a
stochastic model of possible asset movements, a provision that would
be adequate in all but a small proportion of scenarios can be
determined.
____________

23 In most cases the options and guarantees that give a provider of

benefits on future financial events cause for concern are those that are
dependent on future investment returns, or an investment value (yield
or capital value) at some future point in time. Because of the
uncertainty, a stochastic investment model should be used to assess
the provisions necessary for such guarantees.

If future returns exceed a certain level, or if a value or index is above

(or below) a fixed value at some future point, there will be no cost to
the company. But if they are below (or above) that level, there will be a
cost, which increases as returns reduce. Hence a range of future
investment scenarios should be tested.
____________

Page 22 © IFE: 2019 Examinations

 Exclusive use

Example

A unit-linked life assurance policy guarantees to pay a maturity value

equal to the sum of premiums on the chosen maturity date, or the value
of units allocated if greater. At all other times the surrender value is
based on the value of units.

Describe the steps involved in assessing the provision to be made for

the cost of this guarantee.
____________

24 Solution

The steps involved are:

 Choose a stochastic asset model – a complex model gives better
results but takes longer to run.
 Determine assumptions – particularly unit growth rate mean and
volatility.
 Determine consistent deterministic assumptions for mortality and
surrender rates and future expenses.
 Consider dynamic links between assumptions – eg lapse rates to
unit values.
 Choose a time period – probably annual for efficient running.
 Determine appropriate model points for the portfolio.
 The model will project the unit values to maturity, allowing for
future premiums and all decrements.
 This will be done for a large number of randomly generated
investment scenarios – say between 1,000 and 5,000.
 For each scenario and each model point, the projected unit value
will be compared with the guaranteed maturity value, and the cost
for that particular scenario and model point determined.
 The projected costs are discounted to the present, scaled up by
the appropriate factors, and summed across all model points.
 The average across all scenarios is the expected cost of the
guarantee.
 The variability should be assessed by looking at the quartiles and
5th/95th percentiles, when the results are ranked.

© IFE: 2019 Examinations Page 23

 Exclusive use

 For reserving purposes an appropriate ruin probability needs to be

chosen. Perhaps 1 in 100, in which case the reserve is the 99th
percentile.
____________

25 The results from the models depend on the model itself and the values
assigned to the parameters in the model. Models should not be treated
as black boxes, the output of which is assumed to be correct.

The use of a stochastic model goes some way to illustrating the

potential variability of the experience, but the results that it produces
are still dependent on the accuracy of the model and its parameter
values. In the case of a deterministic model, the potential uncertainty
of the results is greater, because fewer scenarios are tested.

The re-running of a model (deterministic or stochastic) with different,

but feasible, parameter values will produce alternative results and
hence help to illustrate the potential deviations. The re-running with a
series of different sets of parameter values, perhaps chosen from a
probability distribution for such values, will help to illustrate the likely
range in which actual experience may lie, perhaps as far as creating a
probability distribution for this experience.

For example, consideration of the effect of a change in the membership

profile of a funded pension scheme may be needed to illustrate the
extent of potential variability in future contributions if the model used
is based on a stable membership profile.
____________

26 There is a possibility of model error if the model developed is not

appropriate for the financial products, schemes, contracts or
transactions being modelled. Checks of goodness of fit will be needed
to assess the suitability of the model, but taking account of expected
changes in experience into the future.

The effect of mis-estimation of parameter values can also be

investigated by carrying out a sensitivity analysis. This involves
assessing the effect on the output of the model of varying each of the
parameter values. When doing this any correlation between different
parameters should be allowed for.
____________

27 In the case of a model used for pricing, the results from the sensitivity
analysis will help to assess the margins that need to be incorporated
into the parameter values.

Page 24 © IFE: 2019 Examinations

 Exclusive use

In the case of models used to assess return on capital and profitability

of existing business, the results will enable the actuary to quantify the
effect of departures from the chosen parameter values when
presenting the results of the model to the company.
____________

28 The statistical risk associated with the parameter values can be

allowed for through the risk element of the risk discount rate. An
alternative would be to use a predetermined discount rate and then
assess the effect on the results of the models of statistical risk.

Where a probability distribution can be assigned to a parameter, it may

be possible to derive the variance of the profit or return on capital
analytically. More generally, a sensitivity analysis, as described above,
can be carried out. Whichever of these two is used, they will again
help in assessing margins or in quantifying the effect of departures
from the chosen parameter values when presenting the results of the
model.
____________

Chapter 18 – Data

29 Organisations often accumulate large amounts of information relating

to individuals as part of their ongoing operations. The increasing use
of technology has now made it possible to collect, store and use very
large amounts of information about individuals in ever more diverse
ways.

Organisations have particular responsibilities when acquiring and

maintaining personal data.
____________

30 Personal data relates to information in respect of an individual where

the individual can be identified, or where the data combined with other
information could allow the individual to be identified.
____________

31 Organisations have an ethical responsibility to deal responsibly with

personal data. In particular, they need to balance the privacy of
individuals with the need of the organisation to make fair and
reasonable use of the personal data in their operations.
____________

32 Many countries have data protection laws to safeguard the rights of

individuals with regard to how organisations can process and maintain
personal data.

© IFE: 2019 Examinations Page 25

 Exclusive use

While the relevant regulations vary by jurisdiction, the objectives and

expected behaviour are often similar. Examples of legislation that are
broadly similar include the Data Protection Act in the UK, Personal
Information Protection and Electronic Documents Act in Canada, and
Personal Data (Privacy) Ordinance in Hong Kong.

However, not all countries have equivalent data protection legislation.

For example, the USA has much less stringent personal data / privacy
laws or regulations than the UK. Organisations need to take extra care
where data is being transferred between countries, even if the purpose
is valid.
____________

An example of common data protection principles is contained in the

Data Protection Act in the UK, which has eight principles that must be
followed when processing personal data.
____________

33 Personal data must:

 be processed fairly and lawfully
 be obtained and processed for specified purposes
 be adequate, relevant and not excessive for the purposes
concerned
 be accurate and, where necessary, kept up to date
 not be kept longer than necessary for the purposes concerned
 be processed in accordance with the individual’s rights under the
Act
 be processed securely
 not be transferred to a country or territory outside the European
Economic Area unless that country or territory ensures an
adequate level of protection.
____________

Page 26 © IFE: 2019 Examinations

 Exclusive use

34 The consequences of non-compliance with the relevant data protection

laws when processing personal data can be significant.

For example:
 Individuals who commit criminal offences may be prosecuted.
 Organisations can be fined for serious breaches. For example, in
the UK, organisations can be fined up to £500,000.
 In addition to prosecution and/or financial penalties, breaching
data protection rules could lead to adverse publicity which can
lead to significant reputational damage for an organisation.
____________

35 The ability to identify the individual to whom the information relates is

crucial to the definition of personal data. For anonymous data
(ie where that individual cannot be identified) the obligations on an
organisation are often considerably less. For example, in the UK
anonymous data does not constitute personal data and the duties and
obligations of the Data Protection Act do not apply.
____________

36 In addition to data protection laws, jurisdictions may also have

competition laws which may also limit the uses to which data can be
put.

For example, the following may be prohibited:

 anti-competitive agreements – eg data could be shared among a
small number of companies to fix prices in a particular market
 abuse of dominant market position – eg imposing unfair trading
terms, such as exclusivity.
____________

37 There can be significant consequences of non-compliance with

competition laws, including fines, awards for damages and
disqualification of company directors.
____________

© IFE: 2019 Examinations Page 27

 Exclusive use

38 Sensitive personal data can include information related to:

 racial or ethnic origin
 political opinions
 religious or other similar beliefs
 membership of trade unions
 physical or mental health condition
 sexual life
 convictions, proceedings and criminal acts.
____________

39 Sensitive personal data is generally subject to much stricter regulation

than ordinary personal data.
____________

40 For example, it may be the case that sensitive personal data can only
be processed when one of the following conditions has been satisfied:
 The data subject has given explicit consent.
 It is required by law for employment purposes.
 It is needed in order to protect the vital interests of the individual
or another person. For example, if an individual with a medical
condition has an accident at work, it would be in the individual’s
vital interest to disclose this condition to medical staff treating the
individual.
 It is needed in connection with the administration of justice or legal
proceedings.
____________

41 The increasing use of technology has now made it possible for the
public and private sector to collect and analyse very large data sets of
information. This is often referred to as ‘big data’.

Big data can be characterised by:

 very large data sets
 data brought together from different sources
 data which can be analysed very quickly – such as in real time.
____________

Page 28 © IFE: 2019 Examinations

 Exclusive use

42 Big data can include personal data (such as data from social media or
loyalty cards), but can include other data (such as climate change
data). If personal data is held by a company, then the company needs
to comply with the relevant data protection rules. Given the large
amount of information that could be held on an individual, privacy
considerations are likely to be a concern for individuals whose data is
held.

Anonymisation can potentially aid big data analytics, as it means that

the information being analysed is no longer considered personal data.
This can assist organisations to carry on research or develop products
and services. It also enables these organisations to give an assurance
to the people whose data was collected that the organisation is not
using data that identifies them for big data analytics.

A key feature of big data is using ‘all’ the data, which contrasts with the
concept of data minimisation in the data protection principles. This
raises questions about whether big data is excessive, while the variety
of data sources often used in big data analytics may also prompt
questions over whether the personal information being used is
relevant. Organisations need to be clear from the outset what they
expect to learn or be able to achieve by processing the data, as well as
satisfying themselves that the data is relevant and not excessive.

Organisations that hold big data also need to be transparent when they
collect data, and explaining how the data will be used is an important
element in complying with data protection principles. The complexity
of big data analytics will not be an acceptable excuse for failing to
obtain consent where it is required.

Regulators expect organisations that hold big data to be proactive in

considering any information security risks posed by big data.

Data governance is becoming increasingly important for holders of big

data. This must take account of data protection and privacy issues.
____________

43 Data governance is the term used to describe the overall management

of the availability, usability, integrity and security of data employed in
an organisation.
____________

© IFE: 2019 Examinations Page 29

 Exclusive use

44 A data governance policy is a documented set of guidelines for

ensuring the proper management of an organisation's data.

A data governance policy will set out guidelines with regards to:
 the specific roles and responsibilities of individuals in the
organisation with regards to data
 how an organisation will capture, analyse and process data
 issues with respect to data security and privacy
 the controls that will be put in place to ensure that the required
data standards are applied
 how the adequacy of the controls will be monitored on an ongoing
basis with respect to data usability, accessibility, integrity and
security.

The data governance policy will also provide a mechanism for ensuring
that the relevant legal and regulatory requirements in relation to data
management are met by the organisation.
____________

45 Organisations that do not have adequate data governance procedures

can be exposed to risks relating to:
 legal and regulatory non-compliance
 inability to rely on data for decision making
 reputational issues
 incurring additional costs (for example fines and legal costs).
____________

A sound data governance policy should therefore provide the

organisation’s stakeholders (staff, management, regulator,
shareholders and policyholders, amongst others) with confidence that
the organisation is dealing appropriately with the data it holds.
____________

Page 30 © IFE: 2019 Examinations

 Exclusive use

46 Where businesses are combined by merger or takeover, one of the key

issues is whether the data for the two businesses should be combined
onto one system and, if so, which.

The saving in overhead costs such as system maintenance and

management is frequently cited as a justification for the transaction. In
practice, the costs of converting the data from one working system to
another are high. New developments are carried out on one system
and the other is left to decline as a legacy system, often requiring
proportionately higher maintenance costs. Thus the aim of cost saving
is often not achieved.

There is a risk in aggregating data sourced from difference systems

and a data governance policy needs to address this risk.
____________

An actuary is faced with a range of possible risks when using data.

____________

47 Examples of possible risks associated with using data are:

 The available data might contain errors or omissions, which could
lead to erroneous results or conclusions.
 There may be insufficient historical data available to estimate
credibly the extent of a risk, and the likelihood of the occurrence of
that risk in future.
 Even where there is sufficient data to estimate credibly future
experience in normal conditions, there may be insufficient data
available to provide a credible estimate of a risk in very adverse
circumstances, which may be necessary for some purposes (eg
estimating the tails of a distribution).
 Where there is insufficient data it may be possible to use data from
other sources (eg industry data, other countries, competitors), but
there is a risk that data from other sources may not be a
sufficiently good proxy for the risk being assessed.
____________

© IFE: 2019 Examinations Page 31

 Exclusive use

48 Further data risks

 Historical data may not be a good reflection of future experience.
This could be due to:
– past abnormal events
– significant random fluctuations
– future trends not being reflected sufficiently in past data
– changes in the way in which past data was recorded
– changes in the balance of any homogeneous groups
underlying the data
– heterogeneity with the group to which the assumptions are to
relate
– the past data may not be sufficiently up to date
– other changes – eg medical changes, social changes,
economic changes etc.
____________

49 Further data risks

 There are risks where an actuary attempts to group data into
broadly homogenous groups. The risks associated with this are:
– the individual data groups may be too small for a credible
analysis
– if data groups are merged so there is sufficient data in each
group to be credible, the combined data set may not be
sufficiently homogeneous.
____________

50 Further data risks

 The available data may not be in a form that is appropriate for the
purpose required.
 A lack of confidence in the available data will reduce the
confidence in an actuary’s conclusions.
 The available data may have been collected for a purpose, which
means that it is not appropriate for a different purpose.
____________

Page 32 © IFE: 2019 Examinations

 Exclusive use

51 In the last few decades, the trading of financial assets has increasingly
been carried out electronically. Advances in computer power,
communication technology and programming capability have offered
new tools for investment decisions, trading execution and risk
management.

Electronic trading has the advantages of increased speed and

efficiency of trading, and can result in lower dealing costs on trades.
In addition, automated trading can potentially facilitate the execution of
complex trading strategies that would not have previously been
possible.
____________

52 Algorithmic trading is a form of automated trading that involves buying

or selling financial securities electronically to capitalise on price
discrepancies for the same stock or asset in different markets. Often
many trades are carried out very quickly to take advantage of
temporary price discrepancies, with the aim of making small profits on
each trade. The trader will use a formula (or algorithm) to decide
whether a financial asset should be bought or sold.

The parameters underlying the algorithm used to determine when

assets would be bought or sold will need to be derived using data from
an appropriate source(s).
____________

53 The following risks are associated with algorithmic trading:

 There could be an error in the algorithm or the data used to
parameterise the model could be wrong, leading to potential losses
on each trade, rather than the expected profits. This is an issue
when a large number of trades could be completed very quickly.
 The algorithm may not operate properly in adverse conditions. For
example, the algorithm could stop trading an asset in turbulent
markets, reducing liquidity of the asset and increasing volatility.
 In very turbulent conditions, trading in individual stocks, or even
entire markets, may be suspended before an algorithmic trade can
be completed.
 The main risk of algorithmic trading is the possible impact on the
financial system. An example of this was a 5%-6% plunge and
rebound in major US equity indices within the span of a few
minutes due to a large number of trades done at erroneous prices
in May 2010. The increasing integration between markets and
asset classes means that a meltdown in one market could impact
other markets and asset classes.

© IFE: 2019 Examinations Page 33

 Exclusive use

54 Algorithmic trading looks at prices of stocks across all markets. An

early development was known as programmed trading, which just
considered automated rules for trading individual stocks on a single
market. It gives a good example of the advantages and disadvantages
of algorithmic trading.
____________

Example

You own 100 shares of a company, with each share currently equal to
£1. The rules in place for buying / selling the stock require you to sell
the stock when it falls in value by 5% from the starting point, and buy
when the stock rises in value by 5% above its low point.

Say the share price falls from £1.00 to £0.93 on a given day and then
bounces back to £1.05 before the end of the day.
____________

55 Your cashflows would be:

 You sell your 100 shares when the price reaches £0.95 per share,
and you realise £95 in cash (100 × £1.00 × 0.95).
 You then use the £95 in cash to rebuy the shares when the share
price reaches £0.9765 (£0.93 × 1.05).
 This gives you 97.28 shares (£95 / £0.9765), which rise in value to
£102.15 (97.28 × £1.05) by the end of the day.
____________

56 Was it worth selling and re-purchasing the shares?

If you had done nothing at all, the value of your shares would have
risen from £100 to £105 (100 × £1.05) by the end of the day. You have
therefore made a small loss of £2.85 (ie £102.15 – £105) by selling and
buying when you did not need to. In practice this loss would have
been more as there would have been other frictional costs, such as
trading costs, when you bought and sold the stock.

However, it is possible that the share price would not have bounced
back, and possibly it could have fallen further. In this case, selling the
stock before it fell further may have been a good strategy. The loss
you made in the example above might be regarded as an acceptable
insurance premium for the protection that the method gives against
larger price falls.

Page 34 © IFE: 2019 Examinations

 Exclusive use

The investor therefore needs to consider whether the potential for loss
outweighs the possible benefits when designing a trading scheme
which is based solely on numbers rather than on any research into the
company involved.
____________

Actuaries use data in all their work. The interaction between the data
requirements for the various tasks actuaries carry out can be complex
and will vary from organisation to organisation. Essentially, however,
for a given type of work the underlying data requirements will normally
be similar.
____________

57 The overriding principle is that the data for all the tasks should be
controlled through one single, integrated data system. However, this
ideal is not always achieved in practice. In a smaller organisation it is
easier to ensure that the data used for different applications are
consistent, because it is likely that the same small group of people will
carry out the applications.
____________

58 For some purposes, data may only be required on a ‘big picture’ basis.
Here, data will be publicly available from published company accounts
and regulatory returns.

Product providers need data relating to the individual risks that they
provide cover for. The quantity and quality of these data are both
important. Without sufficient quantity, data groupings will either be
non-homogeneous or lack credibility. However, even where there are
plenty of data available, poor quality data will mean that any results
produced are not reliable.
____________

59 Problems of data quality and quantity can be a result of poor

management control of data recording or its verification processes, or
due to poor design of the data systems. This may not necessarily be a
reflection on the current management, as good quality data cannot
necessarily be obtained quickly. After implementing a process for
maintaining extensive records, it may take many years for enough data
to be collected for analysis purposes.
____________

The availability of data of good quality and quantity will vary greatly
between organisations and, within organisations, between the different
classes of business.
____________

© IFE: 2019 Examinations Page 35

 Exclusive use

60 When placing a value on liabilities, for health care, life and general
insurers, the prime information source will be the details given on the
proposal form. It is therefore important that it produces relevant and
reliable information for the system.

Questions need to be well designed and unambiguous, so that the

proposer will give the full, correct information and the underwriting
department can process the application readily, adding any coding that
is necessary. In particular the result of any medical or occupational
underwriting will need to be added. For general insurance personal
lines the composition of the final premium from various rating factors
will be important.

This information (together with any subsequent changes) will need to

be held for several purposes, including cross-checking against the
claims information at the time of any claim. This should enable the
automatic checking of the validity of the claim and the updating of the
policy information (eg termination of cover in the event of a total loss
under a general insurance policy or death under a life insurance
policy). The data requirements will depend on the type of benefits
provided.

Another important information source will be the details given on a

claim form. Like the proposal form, it is important that this is designed
with the aim of producing information that can be both analysed
accurately and also transferred easily to the computer system.

As well as data relating to current risks covered, it is important to

retain the history of past policy and claim records.
____________

61 There may be occasions when the actuary does not have full control
over the data available. For example, when valuing benefits under an
employee benefit scheme, the scheme sponsor will usually provide
data on the operation of the scheme and the scheme membership. It
will be particularly important to validate this type of data.

Data will be required to place a value on the benefit entitlements of

individuals. Data will be required in respect of individuals who have an
entitlement to receive a benefit in the future and also individuals who
are currently receiving benefits. The data will need to be sufficiently
detailed to provide all information that is likely to be financially
significant to the level or timing of future benefits.

Page 36 © IFE: 2019 Examinations

 Exclusive use

For example, if a pension is to be provided, the age of the individual

will be significant. However, if a pension were also to be paid to a
spouse after the death of the member, the existence and age of a
spouse of a young member may not be financially significant as the
marital status of the member may change in the future.

Any equivalent data used when previously valuing benefits will be

useful to the actuary as it will enable reconciliations to be performed
that help to indicate the validity of the current data. Accounting data
may also help in this process.

Where reserves are built up for benefits, a balance sheet and income
and expenditure statement may exist. This will provide information
about the total value of the assets held and perhaps information
relating to recent benefit outgo and premium / contribution income.
This information will be useful in verifying other data or in considering
the assumptions to be used. If audited accounts exist, they will enable
greater reliance to be placed on the figures when verifying the data.

To place a value on assets that is reliable and consistent with a value

placed on future benefits, it is necessary to obtain a full listing of the
individual assets held. These individual holdings should then be
checked to determine whether they are permitted or are subject to
valuation restrictions imposed by regulation or legislation.
____________

62 Whether using data provided by their own organisation or a third party,

an actuary will have to make and check certain assertions about that
data. Such assertions include:
 that a liability or asset exists on a given date
 that a liability is held or an asset is owned on a given date
 that when an event is recorded the time of the event and the
associated income or expenditure are allocated to the correct
accounting period
 that data is complete, ie there are no unrecorded liabilities, assets
or events
 that the appropriate value of an asset or liability has been
recorded.
____________

A decision will then have to be made as to how these assertions will be

checked and the level of detail that will be appropriate in checking
them.

© IFE: 2019 Examinations Page 37

 Exclusive use

63 Possible checks could include:

 reconciliation of the total number of members / policies and
changes in membership / policies, using previous data and
movement data
 reconciliation of the total benefit amounts and premiums and
changes in them, using previous data and movement data
 the movement data should be checked against any appropriate
accounting data, especially with regard to benefit payments
 checks should be made for any unusual values, such as
impossible dates of birth, retirement ages or start dates
 consistency between salary-related contributions and in-payment
benefit levels indicated by membership data and the
corresponding figures in the accounts
 consistency between the average sum assured or premium for
each class of business should be sensible, and consistent with the
figure for the previous investigation
 consistency between investment income implied by the asset data
and the corresponding totals in the accounts
 where assets are held by a third party, reconciliation between the
beneficial owner’s and the custodian’s records
 full deed audit for certain assets, such as checking the title deeds
to large real property assets
 consistency between shareholdings at the start and end of the
period, adjusted for sales and purchases, and also bonus issues,
etc
 random spot checks on data for individual members / policies or
assets.
____________

Page 38 © IFE: 2019 Examinations

 Exclusive use

64 The main circumstances where ideal data are not available are
because:
 Data have not been captured at a sufficiently detailed level. For
example, a benefit scheme may not analyse membership by
whether the employee is a clerical or a manual worker. Changes in
the structure of the membership may have a material effect on
scheme benefits such as early death or accident benefits.
Similarly, where life assurance premiums are collected at the door
by an agent who calls, only limited data may be captured on the
insurer’s database.
 There may be insufficient data to provide a credible result. A
provider may have recently launched a new product or branched
out into a new target market. Alternatively the provider may simply
be too small to attach any credibility to its own experience. This is
particularly the case with benefit schemes, where very few
employers will be of sufficient size to have credible experience to
assess mortality rates before retirement.
____________

65 When valuing benefits it may be appropriate to use summarised data

instead of detailed membership data in some circumstances. However,
it should be recognised that the reliability of the values will be reduced,
as full validation of the data will be impossible. Additionally, the
summarised data may miss significant differences between the nature
of benefits that have been grouped together. Summarised data is
therefore only suitable if such inaccuracy is recognised by the users of
the results of the calculations.

It is also unlikely that summarised data could be used to value options

or guarantees that may or may not apply on an individual basis.
____________

66 In some countries there are organisations that collect data from their
member offices and then make available summaries of all the data to
their members. For example, in the UK, the Association of British
Insurers collects and collates a wide variety of insurance data. This
cannot be used in place of policy data to establish provisions for a
particular policy or scheme, but could be used to determine bases or
be used in product pricing.

One of the best examples is the Continuous Mortality Investigation

Bureau of the Institute and Faculty of Actuaries in the UK, which does a
large amount of work on mortality and morbidity statistics. The volume
of data that can be collected from across a whole industry greatly
improves the statistical significance of the resulting analysis.

© IFE: 2019 Examinations Page 39

 Exclusive use

67 An insurer participating in an industry-wide scheme has the prospect

of being able to compare its own experience with that of the industry
as a whole (or that part of it represented by the participating insurers)
with regard to both the overall level and the pattern of the experience
by the categories into which the data are classified. Any significant
differences point to a need for explanation. Since an insurer is likely to
be seeking to expand by attracting business from its competitors, it
may be important to have an indication of the ways in which the
characteristics of the business it is seeking may differ from those of
the business it already has.
____________

68 When using industry-wide data, there is potential for distortions arising

from heterogeneity. This is because the data supplied by different
organisations may not be precisely comparable because:
 companies operate in different geographical or socio-economic
sections of the market
 the policies sold by different companies are not identical
 sales methods are not identical
 the companies will have different practices, eg underwriting or
claim settlement standards
 the nature of the data stored by different companies will not always
be the same
 the coding used for the risk factors may vary from organisation to
organisation.

Other problems with using industry-wide data may be:

 the data will usually be less detailed, or less flexible, than those
available internally
 external data are often much more out of date than internal data
 the data quality will depend on the quality of the data systems of
all of its contributors
 not all organisations contribute, and the organisations that do
contribute are not representative of the market as a whole.
____________

69 It may also be possible to obtain data from a reinsurer or from national

statistics.
____________

Page 40 © IFE: 2019 Examinations

 Exclusive use

70 The main aim of risk classification is to obtain homogeneous data. The

reduction of heterogeneity within the data for a group of risks makes
the experience in each group more stable and characteristic of that
group. Furthermore, it enables the data to be used more appropriately
for projection purposes. This is important when monitoring claims and
mortality experience. Any heterogeneity in data groups will serve to
distort the results and can lead to setting provisions that are too big or
too small and calculating premiums or contributions that are incorrect.

Ideally data to be analysed should be split into homogeneous groups,

for example, by age and gender in a mortality investigation. However,
where data is scarce, such as for numbers of deaths at young ages,
splitting data into homogenous groups may result in data groups that
are too small to enable any credible analysis to be carried out. In such
cases data may need to be combined into groups which are less
homogeneous, but which are large enough to be credible. Whenever
data is to be analysed there needs to be a balance between splitting the
data into homogeneous groups and having sufficient data in each
group to enable a credible analysis to be carried out.

There is also a need to carry out sensitivity testing to check that if the
data are grouped in a different way the same results are obtained.
____________

Chapter 19 – Setting assumptions

71 As in all actuarial work, when setting assumptions it is important to:

 consider the use to which the assumptions will be put
 take care over the choice of the assumptions that will have the
most financial significance
 achieve consistency between the various assumptions
 consider any legislative or regulatory constraints
 consider the needs of the client.
____________

Historical data is likely to be a primary source of data used in

determining assumptions about future experience.
____________

© IFE: 2019 Examinations Page 41

 Exclusive use

72 Historical data can be helpful when choosing demographic

assumptions. For example, historical levels of mortality in a country,
industry or company may help with the choice of assumptions for the
number of individuals who will survive to receive pensions, or for the
extent to which contingent benefits will be payable. In many countries
this data will have been analysed and used to produce a graduated
decrement table. Past data can also be used to project future
improvements in mortality.

Similarly, past data can be used when determining the probability of

individuals leaving employment, becoming ill, retiring, being married or
other significant life events.

In determining an assumption for future investment returns, past data

on dividend yields on equities and on the total returns on relevant
classes of investment may be useful. Where dividends are linked to an
inflation index, past data on that index may be useful.

Past data on salary levels in a particular country, industry or company

may be useful when making an assumption about future levels of
salary growth. The history of an inflation index may also be useful in
determining an assumption for future benefit growth that is linked
either fully or partially to that inflation index.
____________

Current data may also be of use when determining assumptions.

____________

73 The relationship between current yields for fixed-interest and index-

linked bonds may provide some indication of the market’s view of
future levels of the inflation index to which the bonds are related.

Policy statements by governments or controlling banks may also be

useful when making assumptions about economic factors.

A scheme sponsor may be able to provide information on planned

future salary increases or likely future rates of withdrawal.
____________

74 In some instances, assumptions may be defined in regulations or

legislation.
____________

Page 42 © IFE: 2019 Examinations

 Exclusive use

75 It is unlikely to be sensible to take an average rate from past data and

assume that it will be appropriate for projecting future experience.
Past data does not provide the answer as to what will happen in the
future, but simply provides information that can be considered when
determining the most likely future experience.

The social and economic conditions are likely to have changed over
any period of history. The actuary therefore needs to consider the
conditions that will apply in the future period to which the projections
will relate and how those conditions will lead to differences from the
past data that is being used.
____________

76 The relevance of past data to future projections must also be balanced

against the need for sufficient data for its analysis to be statistically
credible. In making a judgement about future experience this conflict
between credibility and relevance must be managed.
____________

77 When using past data it is necessary to consider how to deal with:

 abnormal fluctuations
 changes in the experience with time
 random fluctuations
 changes in the way in which the data was recorded
 potential errors in the data
 changes in the mix of homogeneous groups within the past data
 changes in the mix of homogeneous groups to which the
assumptions apply.
____________

78 Economic data fluctuates with changes in economic and fiscal policy

as well as with the general economic cycle. Past data for investment
returns, salary levels and dividend yields in most countries fluctuate
significantly over an extended time-frame. It could be thought that
economic and fiscal changes mean that most past data are irrelevant
and so only data that relate to a period after any recent significant
change can be used. However, this would reduce the credibility of the
data and increase the effect of any random fluctuation. It is necessary
to use the earlier data and to try to strip out the fluctuations that relate
to economic and fiscal conditions that differ from those that currently
exist.
____________

© IFE: 2019 Examinations Page 43

 Exclusive use

79 Past levels of an index to measure price inflation usually fluctuate

significantly and are often a useful indicator of the economic
conditions that existed. They are therefore unlikely to be very useful in
determining an assumption for future levels of inflation. Consequently,
current index values may be a better guide to future levels of inflation.
For example, government projections and the ‘risk-free’ real returns
indicated by the current yields on long-term index-linked bonds could
be used.
____________

80 Past data for price inflation can be very useful in determining other
economic assumptions, as conversion of past economic data into real
terms will often remove much of the fluctuation.
____________

Making any further adjustment for economic or fiscal changes is

difficult to do other than subjectively.
____________

81 Dividend levels could be adjusted to allow for changes in taxation

applying to those dividends. However, an explicit adjustment may be
spurious, as there may be changes to the taxation of companies or
individuals that have a more significant effect.
____________

82 Much of the demographic data will also be affected by economic

changes. Again, explicit adjustment is difficult and so judgement and
analysis of fluctuations and trends will be important.

Mortality data is mainly affected by medical advances. Past data

should be considered with this in mind. This is likely to result in
significant emphasis being placed on the most recent data with
consideration of past trends and their underlying reasons being
important in determining the extent of future change.
____________

83 It is important that the past data used is relevant to the group of

individuals about whom assumptions are to be made. Levels of salary
growth and mortality, for example, usually differ by type of employment
or social class. Ideally the past data would be split into homogeneous
groups to reflect such differences. In practice the information
necessary to split the data reliably is unlikely to be available, and
splitting the data would result in a significant reduction in credibility.
Therefore, past data will usually need to be adjusted in a subjective
manner to allow for differences in the characteristics of the individuals
concerned.

Page 44 © IFE: 2019 Examinations

 Exclusive use

In adjusting past data it is important to recognise that the past data

may give false results due to changes in the balance of homogeneous
groups over time. For example, past levels of salary growth may
reflect a change in the overall composition of a workforce (for example
production line workers being replaced by mechanisation) rather than
just the changes in real salary levels for individuals.
____________

84 Over time, statistics produced by the State or data recorded by

companies may change. Such changes distort the past data and could
lead to inappropriate assumptions unless these changes are
recognised.

Data errors will also cause distortions but may not be as easy to
recognise as changes in the ways of recording the data. Generally, the
management and verification of data recorded by companies has
improved significantly as the capability of computers has improved.
Older data may carry a greater risk of data error, perhaps to an extent
that outweighs the usefulness of having more data.
____________

One-off impacts in the past data will also need to be considered to

ensure that the assumptions are valid.
____________

85 For example, significant returns in one year on a specific asset could

be because of government intervention. High numbers of deaths could
be due to an epidemic meaning that mortality experience for that year
is unusually high.
____________

86 In some countries past data has been analysed on a national or

industry level. The most common data for such analyses relates to
mortality and morbidity. Countries may analyse the whole population’s
experience based on censuses. One disadvantage of census data is
that it includes all lives, and not just the restricted population that
effect insurance contracts. Thus, census data generally includes lower
socio-economic groups that distort the experience of lives effecting
insurance contracts.
____________

© IFE: 2019 Examinations Page 45

 Exclusive use

87 As an example, in the UK the Institute and Faculty of Actuaries has

established the Continuous Mortality Investigation Bureau (CMI), which
produces mortality tables based on experience provided by
participating insurance companies. The CMI Bureau conducts
statistical analysis of the data including, for example, projecting
mortality improvements for the experience of annuitants.
____________

88 When using standard tables, the same considerations are needed as

when using the company’s own past experience data:
 whether the data is relevant to the intended population at which
the product is marketed
 whether adjustments need to be made to the data to reflect
continuation of past historical trends.
____________

89 When considering the assumptions to use to project future experience,

the actuary needs to consider the purpose for which the assumptions
are to be used and the significance of each assumption in the overall
result. This helps assess the degree of accuracy required and hence
the extent to which it is necessary to try to remove distortions from
data. It also helps judge whether the assumption should reflect the
best estimate of the future experience or whether it is appropriate to
reflect any uncertainty about future experience by an overstatement or
an understatement within the assumption.

Where assumptions are used to place a capital value on future

cashflows, it is usually unnecessary to make a judgement about the
accuracy of each assumption. Instead it is necessary to determine that
the overall value resulting from the combination of assumptions is
appropriate. However, where the individual cashflows are important, it
may be necessary for the accuracy of each assumption to be assessed.

Consideration of the potential financial significance of errors in the

assumptions also helps assess the degree of accuracy required, the
extent of margins necessary or the level of risk being taken.
____________

90 For example, when acting as expert witness to determine a fair

compensation settlement between two parties, it is important that the
assumptions used are the actuary’s best estimate of the future
experience. Under- or over-statement will give one party a direct
financial advantage at the expense of the other. Consequently each
party will have a preference for which side of ‘best estimate’ they
would like to see each assumption.

Page 46 © IFE: 2019 Examinations

 Exclusive use

91 It is necessary to be aware of implicit assumptions within a model, and

consider the effects of these. For example, the funding method for an
occupational pension scheme may assume that:
 new members continue to join or new policies continue to be
written and therefore the age / sex distribution of a population will
be maintained
 no new entrants will join or no new policies will be written and so
the existing population should be treated as a closed group.
____________

92 The assumptions will be estimates of the expected values for the

parameters. Where a cashflow model is being used to price a product,
the risk to the provider from adverse future experience could be
allowed for by:
 adjusting the risk element of the risk discount rate
 using a stochastic discount rate
 applying margins to the expected values.
____________

93 In pricing a product, a profit requirement will need to be incorporated.

This is because it is reasonable to suppose that the owners of the
provider decide where to invest by comparing the returns offered by
different projects, relative to the risks that are run.
____________

94 Not all products are equally risky. The provider should view itself as
an investor like any other when it considers the risks of a new product,
as in the long run the profits emerging from the company are the
profits emerging from the products that it sells.

A change in the mix of business, for example away from old and safe
contracts towards new and innovative contracts, would change the
market’s evaluation of the provider’s risks.

© IFE: 2019 Examinations Page 47

 Exclusive use

The following features can increase the risks in a product design:

 lack of historical data
 high guarantees
 policyholder options
 overhead costs
 complexity of design
 untested market.
____________

It is not easy to assess these risks, and it is even harder to say what
effect they should have on the risk discount rate.
____________

95 A profit criterion is often a single figure that tries to summarise the

relative efficiency of contracts. By applying a profit criterion to
different contracts and then ranking the results in order, it may be
possible to determine which contracts makes most efficient use of a
company’s capital.

The methods of quantifying profitability include:

 net present value
 internal rate of return
 discounted payback period.

For example, a possible profit criterion for an insurer is that the net
present value of profits emerging from each of its product lines is a
pre-determined proportion of the distribution costs. Such a criterion
reduces the bias towards products with high commission rewards in
the distribution system.
____________

Chapter 39 – Glossary

96 Best estimate

An actuarial assumption which the actuary believes has an equal

probability of under or overstating the future experience (ie the median
of the distribution of future experience.)

Page 48 © IFE: 2019 Examinations

 Exclusive use

Internal rate of return

The discount rate at which the Net Present Value of a series of

cashflows is zero.

Profit test

A profit test is a technique involving consideration of the cashflows

arising under a contract to assess the expected profitability of that
contract. It can be used to determine the premium or the level of
charges under a contract.

Risk discount rate

A risk discount rate is a rate at which future uncertain cashflows might

be discounted. It typically arises when carrying out a discounted
cashflow assessment of value of a project. It represents the risk-free
rate of return that the providers of capital demand plus an amount to
allow for the risk that the profits may not emerge as expected from the
project.
____________

© IFE: 2019 Examinations Page 49

 Exclusive use

PAST EXAM QUESTIONS

This section contains the exam questions from 2008 to 2017 that are related
to the topics covered in this booklet.

Outline solutions are given later in this booklet. These give a summary of
the relevant points that should have been covered in your solution. Fuller
explanations to each point should be given in the exam but you should be
able to use these solutions to check that you would have covered all of the
main points. Further information may be available in the Examiners’ Report,
ASET or Course Notes.

1 Subject CA1 Paper 1 April 2008 Question 2 (part)

List the items that an insurance company would consider when determining
the assumption for future expense inflation in its financial modelling. [2]

2 Subject CA1 Paper 1 September 2008 Question 4 (part)

An individual aged 50 has been told that he will be made redundant from his
job in one month’s time. He has been offered an immediate early retirement
pension from his employer’s pension scheme and a redundancy cash
payment. An actuary has been asked to give advice on how the individual
could use the redundancy cash payment.

Discuss the features of the cashflow projections the actuary would carry out
when giving advice. [8]

3 Subject CA1 Paper 1 September 2008 Question 5 (part)

The company intends to carry out an analysis of its expense experience for
the purpose of product pricing.

(i) Describe the factors the company should consider when using the
results of this analysis. [6]

Rather than carrying out a new, full analysis, the directors are considering
using the results of the previous analysis. They suggest taking the previous
expense figures and adjusting them to allow for inflation.

(ii) Discuss this approach. [6]

[Total 12]

Page 50 © IFE: 2019 Examinations

 Exclusive use

4 Subject CA1 Paper 2 September 2008 Question 5 (part)

A large general insurance company writes warranty business for

construction companies that build housing. The warranties are for a period
of twelve years and cover major damage caused by building defects in new
homes. The company has been writing this business for the last ten years.

Outline the main data issues that will need to be considered when
calculating a theoretical risk premium for this business. [2]

5 Subject CA1 Paper 1 April 2009 Question 3 (part)

A life insurance company launched a single premium unit-linked bond

product five years ago. The terms of the contract were as follows:
 Annual Management Charge of 1% of the accumulated fund each year
 early exit charges to apply in first year at 3% of the accumulated fund,
second year at 2% and in third year at 1%
 guaranteed return of premium on death.

The policyholders have the choice of three funds to invest in:

 an equity fund
 a fixed interest fund
 a fund that guarantees at least the return of premium on every
anniversary of the policy.

Describe the approach that the company would have used to project its
expected profits when it launched the product. [6]

6 Subject CA1 Paper 2 April 2009 Question 3

An actuary has been asked to carry out a small, one-off investigation for a
client, which will require the use of a model.

Discuss the factors the actuary should take into account when considering
how to approach the development of this model and choosing the source of
the model. [8]

© IFE: 2019 Examinations Page 51

 Exclusive use

7 Subject CA1 Paper 2 April 2009 Question 7 (part)

Company X is a subsidiary of a large multinational group, Group Y. Group Y

has an extensive range of businesses in the country in which X operates.
X is going to be demerged from Y, and the demerger agreement requires X
to set up pension arrangements for its employees, which should be broadly
similar to the defined benefit pension scheme operated by Y. X has very
little in-house resource or expertise for pension operations.

Two years after the scheme was established, an actuarial investigation is

being carried out into the funding position of the scheme, and the
surplus / deficit is being calculated.

(i) Outline the data that is needed for this investigation. [6]

(ii) Outline the checks that would be performed on the data. [4]
[Total 10]

8 Subject CA1 Paper 1 September 2009 Question 4 (part)

A life insurance company in a certain country has been selling term

assurances directly to policyholders via the internet for five years. The
company is about to conduct an analysis of the mortality experience on this
portfolio.

This analysis shows that the mortality experience on this portfolio is much
lighter than what was allowed for in the pricing basis.

(i) Discuss possible reasons for this. [3]

Currently, the company has different premium rates for males and females.
The government of this country is about to introduce legislation banning
gender discrimination in the provision of goods and services. There are no
exceptions for insurance.

(ii) Discuss the effect this legislation will have on the approach that this
company uses to set premium rates. [5]
[Total 8]

Page 52 © IFE: 2019 Examinations

 Exclusive use

9 Subject CA1 Paper 2 September 2009 Question 7 (part)

A company is considering purchasing a large office building located in a

large city and is assessing the options available to manage the risks
associated with the purchase.

(i) Discuss the requirements of a model to be used to assess these risks. [9]

(ii) Outline the steps required to evaluate these risks using a model. [5]
[Total 14]

10 Subject CA1 Paper 1 April 2010 Question 1

Outline the key requirements of a model that will be used to assess the
variability of an outcome that depends on uncertain future events. [5]

11 Subject CA1 Paper 1 April 2010 Question 6 (part)

A mutual life insurance company has a large number of policyholders and a

significant undistributed surplus.

(i) List the data required to calculate the value of the provisions needed for
the in-force policies. [4]

(ii) Outline why the data used to determine these provisions would be
grouped. [2]

(iii) Explain the considerations that should be taken into account when
validating the grouping of the data. [4]
[Total 10]

12 Subject CA1 Paper 1 April 2011 Question 6 (part)

A company which writes general insurance business is developing a new

line of insurance policies aimed at university students.

Outline how the assumptions used to price these policies may differ from
those used for the company’s existing business. [9]

© IFE: 2019 Examinations Page 53

 Exclusive use

13 Subject CA1 Paper 2 April 2011 Question 2 (part)

In conjunction with a marketing promotion, a life insurance company has

decided to offer a free one-year term assurance product to mothers of
children born during a particular promotional period. The sum assured is
£10,000 payable on the death of the mother if this occurs before the child’s
first birthday.

Application forms are included in a pack given to prospective new mothers.

Outline the assumptions that the life insurance company would need to
make in order to evaluate the size of the risk being taken on. [2]

14 Subject CA1 Paper 2 April 2011 Question 6 (part)

A large financial institution has a significant credit card business covering a

wide range of personal customers in its domestic market. Different
customers are charged different interest rates appropriate to their
circumstances. Following the changes in short-term interest rates, the
institution is reviewing the interest rates it charges to its credit card
customers.

Describe the process the credit card company would use to establish a
revised interest rate charging structure. [6]

15 Subject CA1 Paper 2 April 2011 Question 7 (part)

A company which previously sold with-profits life assurance directly through

employed agents is now closed to new business. The company is collecting
data in order to investigate its mortality experience.

(i) Explain why ideal data for this purpose may not be available. [2]

(ii) Discuss how the company can use the limited data it has collected.
[4]
[Total 6]

Page 54 © IFE: 2019 Examinations

 Exclusive use

16 Subject CA1 Paper 1 September 2011 Question 6

In three years’ time, a small provincial town will be celebrating the 750th
anniversary of its founding. The organising committee is planning large and
expensive celebrations to be held over a weekend during the anniversary
year. The committee is worried about the festivities being disrupted or
ruined by bad weather.

The part of the country where the town is situated has notoriously
unpredictable weather that can cover the full range of possible conditions
from mild to extreme.

Data exists covering local, regional and national weather records over the
last few centuries. However, the more local the data is, the less complete
and accurate it is.

A senior lecturer at the town’s college is proposing to use this data in a

model that will predict the best date for this event.

(i) Suggest the criteria that could be used to assess whether the weather
on a particular weekend is likely to disrupt or ruin the event. [3]

(ii) Discuss the extent to which the available data may be appropriate for
the prediction model. [9]

(iii) Outline other considerations that could influence the committee’s choice
of the date for the celebratory weekend. [2]

(iv) Set out the difficulties that the committee could face if it tried to fully
transfer its financial risks on this project to a third party. [4]
[Total 18]

© IFE: 2019 Examinations Page 55

 Exclusive use

17 Subject CA1 Paper 1 April 2012 Question 4 (part)

(i) Outline why companies in the financial services sector have moved to
using more complex models when designing and monitoring products.
[3]

A large financial institution is about to launch an investment fund designed

for experienced investors. The fund will invest directly in a large number of
markets and investment classes.

The institution also intends to offer an insurance policy linked to the new
fund. The insurance policy will provide protection against large falls in the
value of an investor’s holding in the new fund.

The institution will know what assets the fund will invest in. Hence the team
responsible for designing and managing the fund believe that the insurance
policy premiums can be invested in derivatives so as to hedge the risks
faced by investors in the fund.

Therefore, by charging an appropriate premium, the design team says that

there should be little risk to the institution in offering the insurance policy and
guaranteed profits could be made.

The specific investments for the insurance policy will need to be determined
by the design team using a complex model.

(ii) Explain why the need to use a complex model may make it difficult for
the directors of the institution to understand the actual financial risks
associated with the investment of the insurance policy premium. [5]
[Total 8]

18 Subject CA1 Paper 1 April 2012 Question 5 (part)

A country has a large horseracing industry. Horses generally start racing

between the ages of two and five. Most horses are retired from racing when
they reach the age of thirteen. A very small number of horses are retired at
younger ages for breeding. Revenues from breeding fees can be very high
but are very volatile.

It has been proposed that a proportion of all breeding fees should be used to
provide contributions to a general fund that will be invested in order to
provide benefits to look after horses in retirement.

All horses actively involved in racing would be covered by the fund.

Page 56 © IFE: 2019 Examinations

 Exclusive use

Benefits would normally only be payable when horses retire from racing
aged thirteen.

Benefits would only be available in respect of retirement before age thirteen

if the horse is certified as injured by a registered vet.

No benefits would be payable on the death of the horse either before or after
retirement.

Outline the demographic information that would need to be considered when

determining the cost of any retirement benefits. [5]

19 Subject CA1 Paper 1 April 2012 Question 7 (part)

The Wetsock music festival has been running for 25 years. The festival is
open air and based in a remote location. Popularity has increased in the last
five years with a large number of visitors choosing to camp at the site over
the weekend. Although the acts at the festival are known in advance, late
changes to scheduled performances are common.

There are two kinds of ticket and both require wristbands to be worn for the
duration of the festival:

Silver band

Provides access to the festival and campsite.

Gold band

Provides access to the festival, campsite and VIP areas. In addition mobile
text message alerts are provided when scheduled performances are
confirmed or altered.

Next year, an insurance company will sponsor the festival. The insurance
company has negotiated with the organisers to provide free insurance to
gold band ticket holders. The insurance company will receive a share of the
gold band ticket price in return for providing the insurance.

The cover the insurance company decides to offer to gold band holders
includes two key benefits:
 Mobile phone cover: a compensation payment if a visitor has their
mobile phone stolen or damaged at the festival; and
 Cancellation cover: a full refund of the ticket price if the entire festival is
cancelled due to poor weather.

© IFE: 2019 Examinations Page 57

 Exclusive use

(i) Discuss the factors that would need to be considered when determining
the cost of providing the mobile phone cover at this particular festival.
Your description should include possible sources of data. [8]

(ii) Explain why the claims experience for the cancellation cover is likely to
be markedly different and more difficult to predict compared to that for
the mobile phone cover. [4]
[Total 12]

20 Subject CA1 Paper 2 April 2012 Question 5 (part)

A medium-sized manufacturing and distribution company currently has no

formal provisions for sick pay. The written contractual employment terms
are that if an employee is unable to work due to sickness the employee will
not be paid by the employer.

The company operates an informal, discretionary policy where, depending

on circumstances and on the decision of their immediate manager, some
employees do receive some sick pay. The level and duration of such
payments varies and can be changed or stopped with minimal notice.

The company is now proposing to introduce formal sick pay arrangements.

Employees who are absent due to sickness will receive full pay from the
company for five working days without the need to produce evidence of
sickness.

After that time, if certification of sickness from a doctor is provided, payments

will be made as a percentage of full pay and the percentage will be as
follows:
 first three months 100%
 next three months 50%.

Thereafter, payments will cease and no discretionary payments will be made

under any circumstances.

(i) Explain why the company’s recent sickness experience may not be
suitable for estimating future sickness experience and hence payments
arising under the new arrangements. [9]

(ii) Set out difficulties the company may face were it to try to use any other
sources of experience data for sickness rates. [3]
[Total 12]

Page 58 © IFE: 2019 Examinations

 Exclusive use

21 Subject CA1 Paper 1 September 2012 Question 3

A subsidiary of a multinational general insurance group is planning on

launching a new car insurance product covering accident risks (ie not theft
risk). This subsidiary writes business in only one small country and has no
experience of writing car insurance risks.

The subsidiary has decided to use combined data from other companies in
the multinational group to set the assumptions used to calculate premiums.

(i) Describe how useful this data would be in determining these premium
rate assumptions. [4]

The government of the subsidiary’s country has banned the use of both age
and gender as rating factors for all general insurance products.

(ii) Discuss other rating factors that could be used to ensure that premiums
charged by the subsidiary accurately reflect the risks associated with
each policyholder. [7]
[Total 11]

22 Subject CA1 Paper 1 April 2013 Question 2

Describe the requirements which should be satisfied when planning,

designing and building a model. [8]

23 Subject CA1 Paper 1 April 2013 Question 4 (part)

A large insurance company is reviewing the profitability of its pet insurance

product. The pet insurance pays out a fixed amount of money on the death
of the insured pet and will also pay 75% of any veterinary bills as long as the
treatment provided is on the approved list set out in the policy document. A
number of exclusions and restrictions apply.

Outline how to project the expected profits from this product. [7]

© IFE: 2019 Examinations Page 59

 Exclusive use

24 Subject CA1 Paper 1 April 2013 Question 6 (part)

An insurance company has introduced a new term assurance product. At

outset, the policyholder selects an initial term of 10, 15, 20 or 30 years
where the premium will be level, thereafter the policy continues without
further underwriting with increasing yearly reviewable premiums. The policy
ceases at age 95. The regulator requires that the initial policy documents
specify guaranteed maximum premium rates for the yearly reviewable
premiums.

Describe the sources of data that could be used to price the contract. [4]

25 Subject CA1 Paper 2 April 2013 Question 5 (part)

A large insurance company, which has been selling various personal lines
products for many years, is now considering selling travel insurance.

Outline the items of data that the insurance company would need in order to
price this new travel insurance product. [6]

26 Subject CA1 Paper 2 April 2013 Question 6 (part)

(i) List the main uses of data in actuarial work. [4]

A defined benefit pension scheme is closed to future benefit accrual. Its

latest annual valuation has just been completed and this shows that the
scheme’s surplus has increased to ten times its previous level.

(ii) Outline the checks on the data that should be made to ensure the
surplus is correct. [7]
[Total 11]

27 Subject CA1 Paper 1 September 2013 Question 4

(i) List four ways in which scenarios could be derived for use in a capital
model. [2]

(ii) Outline the issues that an insurance company should consider when
building a capital model. [9]
[Total 11]

Page 60 © IFE: 2019 Examinations

 Exclusive use

28 Subject CA1 Paper 2 April 2014 Question 3

A large life insurance company has just completed its internal lapse
experience analysis. The lapse rates are lower than for competitors with
similar products (as shown in industry results). A director of the company
has suggested that the internal analysis may not be correct and that the
industry average lapse rates should be used instead.

(i) Discuss the reasons for the suggestion to use industry lapse experience
and why this may not necessarily be a good idea. [6]

(ii) Describe the validation that should be carried out on the internal
experience analysis. [5]
[Total 11]

29 Subject CA1 Paper 2 September 2014 Question 3 (part)

A small general insurance company currently sells only one line of business.
The company’s strategic plan is to grow rapidly by acquiring other insurance
companies.

The company’s owners have identified a target company, and now wish to
determine an offer price for the target.

Outline the requirements of an actuarial model to place a value on the target

company. [7]

30 Subject CA1 Paper 1 April 2015 Question 7 (part)

A mutual life insurance company sells only annuity business.

Describe the basic features of a model which projects the company’s

solvency position over the next 20 years. [7]

31 Subject CA1 Paper 2 April 2015 Question 3

A large bank is about to launch a new product designed for sophisticated

investors. The product will involve direct investment in a number of markets,
together with associated derivatives.

The bank also intends to offer an insurance policy linked to the new product
so that investors who take out insurance will receive protection against large
falls in the value of their holding in the new product.

© IFE: 2019 Examinations Page 61

 Exclusive use

When setting the premium rates for the insurance policy, the bank has used
advanced stochastic techniques to model a range of financial outcomes.
This model is only accessible to and understood by a small specialist team
within the bank and is therefore treated as a ‘black box’ by all others in the
bank.

The results of the model suggest that the bank will be able to invest the
insurance premiums in such a way that the invested premiums will be almost
certain to meet any claims. The Chairman of the bank interprets this to
mean that this insurance policy can, in effect, achieve guaranteed profits.

Discuss why it may not be appropriate for the Chairman to accept the results
of the model in this way. [9]

32 Subject CA1 Paper 2 April 2015 Question 7 (part)

An insurance company is to provide insurance for passengers travelling

through a new city airport which opened six months ago.

The premium for the insurance policy will be paid by the airport at no cost to
passengers. This is an incentive to encourage passengers to travel through
this airport, rather than travelling through nearby alternative airports which
have been established for some time.

The airport has considered the risks which could be covered by the
insurance policy and decided that the following risks will be covered:

1. Indemnity for the loss of, or damage to, luggage for every
passenger, if it occurs at the airport, up to a maximum payout of
$10,000 for each passenger.

2. A payout of up to $100,000 on death or injury if there is an aircraft

‘incident’ on flights arriving at, or departing from, the airport.

Describe possible sources of data, and their limitations, for the risks of
luggage being lost or damaged. [6]

Page 62 © IFE: 2019 Examinations

 Exclusive use

33 Subject CA1 Paper 1 September 2015 Question 5 (part)

(i) State the objectives of and requirements for building a model. [6]

Helen believes that she can generate sufficient profits from betting on the
results of sporting events to pay for her day to day living expenses.

She has identified that a model needs to be built to analyse past data and
use this to predict future results.

(ii) Describe the decisions that need to be made when determining the
modelling approach to be taken to assess these future results. [6]
[Total 12]

34 Subject CA1 Paper 1 September 2015 Question 7

(i) List the main uses of data by a financial institution. [4]

A funeral director has provided funeral services in a large city for a number
of years. It provides these services through the sale of funeral cost plans,
which it only offers directly to individuals.

Under the funeral cost plans individuals pay regular premiums to the funeral
director over a period of ten years, or until death if sooner. From the time
the funeral cost plan is taken out it guarantees that whenever the individual
dies the funeral services will be paid for.

If an individual stops paying premiums while still alive before the ten years is
up then the funeral cost plan is cancelled. No payment is made to the
individual if the funeral cost plan is cancelled.

(ii) Describe the data that will be required to set premiums for these funeral
cost plans. [10]

(iii) Compare the data requirements for an insurance company offering

whole life assurance policies nationally with those for these funeral cost
plans. [8]
[Total 22]

© IFE: 2019 Examinations Page 63

 Exclusive use

35 Subject CA1 Paper 2 September 2015 Question 6 (part)

An organisation sponsors a defined benefit pension scheme for its

employees. Employees will receive a pension based on final salary and
length of service on retirement at age 65 or on earlier retirement on the
grounds of ill health.

If they leave the organisation before retirement they will receive a deferred
pension (based on salary and length of service at date of leaving) payable
on reaching age 65. If they die whilst they are still employed by the
organisation, a death benefit of four times salary will be paid as a cash sum.

The organisation has asked an actuary to investigate the estimated value of

these benefits if the scheme is maintained in its current form.

(i) List the assumptions the actuary will need to make when carrying out
the investigation. [4]

(ii) Discuss the suitability of the historical data likely to be available for
determining each of the assumptions in (i). [10]
[Total 14]

36 Subject CA1 Paper 1 April 2016 Question 5 (part)

An insurance company writes regular premium term assurance policies

which are used in conjunction with mortgage loans.

Outline why the assumptions used for mortality rates will be important in
ensuring that the policies are correctly priced. [4]

Page 64 © IFE: 2019 Examinations

 Exclusive use

37 Subject CA1 Paper 1 September 2016 Question 5

A life insurance company sells annuity business. It has been investigating

its mortality experience.

The investigation has shown a significantly higher number of deaths over the
last three years relative to those expected under the existing mortality
assumptions.

The investigation has also concluded that the higher number of deaths has
been due to smoking related illnesses.

In light of this experience, the insurance company is considering adjusting

the mortality assumptions it uses to price its annuity business.

(i) Discuss the data limitations that the insurance company will need to
consider when deciding whether the results of the investigation are
suitable for adjusting the mortality assumptions used to price its annuity
business. [8]

Approximately 20% of the business written over the last three years has
been for individuals taking out policies with very large annuities.

(ii) Discuss the data sources that could be used to set mortality
assumptions for these individuals. [5]

(iii) Suggest, with reasons, whether the mortality assumptions for this ‘large
annuity’ subsection of the business should differ from those used for
‘standard’ annuity business. [3]
[Total 16]

38 Subject CA1 Paper 2 September 2016 Question 2 (part)

A defined benefit pension scheme provides, as the default option, a level

pension to each member with a 50% spouse’s pension on the death of the
member.

However, the scheme allows members reaching retirement to choose any

particular rate of payment escalation and any particular percentage spouse’s
pension.

The amount of the amended level of pension is then determined by the

scheme applying adjustment factors to the default level of pension.

© IFE: 2019 Examinations Page 65

 Exclusive use

An actuary is advising the trustees of the pension scheme on what

adjustment factors should be used, so as to minimise additional risks to the
funding of the scheme under this arrangement.

Explain how the actuary should set the adjustment factors. [7]

39 Subject CA1 Paper 2 September 2016 Question 7 (part)

A life insurance company is performing a supervisory valuation of its many

lines of business.

Most material lines of business are valued using an industry-standard

actuarial modelling system, called ELL, purchased from an external actuarial
consultancy. Some material lines of business are, however, still valued
using spreadsheets.

ELL is the company’s strategic preferred system for all actuarial calculations
across its operations. It is run and controlled by the company’s central
actuarial team, and receives data feeds directly from the company’s
administration systems.

Spreadsheets are controlled by the valuation team, separate from the central
actuarial team, and rely on data supplied manually by the relevant
administration teams.

(i) Discuss the advantages and disadvantages to the company of

performing its supervisory valuation using ELL compared with the
spreadsheets. [13]

One of the material lines of business which is still valued on a spreadsheet is

required to be valued on ELL instead, in time for the next supervisory
valuation. Relevant project conversion teams will be put in place to plan and
implement this conversion. Implementation can occur only if the valuation
team’s actuary provides written approval that the revised system is fit for the
purpose of the next supervisory valuation. The valuation team’s actuary is
not part of the project conversion teams.

(ii) Discuss the steps and investigations that the valuation team’s actuary
should perform before providing written approval of the revised system.
[8]
[Total 21]

Page 66 © IFE: 2019 Examinations

 Exclusive use

40 Subject CA1 Paper 1 April 2017 Question 4 (part)

A general insurance company provides liability insurance for hairdressers.

The number of policies and premium income for this business has been
increasing each year. The financial results for this business, however, show
that profits are reducing each year, compared to what was expected.

Describe how the expected profits for this business could be projected in
future years. [4]

41 Subject CA1 Paper 1 April 2017 Question 6 (part)

A large financial institution currently has no operations in the insurance

sector.

It has decided to expand into this market by bringing out a new whole life
product. This product will provide a fixed cash sum intended to cover the
policyholder’s funeral expenses and other related costs.

The product will be marketed as simple and low cost for policyholders.

The product will be advertised in newspapers and on television by an

appropriate celebrity.

There is a guarantee that every applicant between the ages of 50 and 75 will
be accepted subject to only minimal underwriting.

(i) Outline the practical problems that may arise due to the lack of data
held by the institution. [3]

(ii) Describe the commercial implications of these data problems and of the
other features of the policy. [6]
[Total 9]

42 Subject CA1 Paper 2 April 2017 Question 4

The regulatory regime for employer-sponsored defined benefit pension

schemes in a developed country allows the trustees of a scheme to set their
own funding level.

(i) List six main assumptions that will be needed. [3]

(ii) Discuss why different schemes might use different assumptions in

setting their funding level. [11]
[Total 14]

© IFE: 2019 Examinations Page 67

 Exclusive use

43 Subject CA1 Paper 1 September 2017 Question 3

(i) List six areas of actuarial work where data would be required. [3]

(ii) Describe how actuaries can mitigate or make allowances for the
problems that may arise from poor quality data. [8]
[Total 11]

44 Subject CA1 Paper 1 September 2017 Question 6 (part)

There has been a distinctive pattern to mortality rates by age in a country for
many years.

Male mortality rates from ages five to seventeen are low and similar.
Thereafter, they rise steeply, peaking at age twenty five. They then fall
steadily to age thirty where the mortality rate is similar to that at age
nineteen.

Female mortality rates from ages five to thirty are lower than male rates and
generally rise steadily with age without the peak at age twenty five prevalent
in male mortality rates.

Recent anecdotal mortality evidence from the last three years appears to
imply that female mortality rates are changing and now show a similar
pattern to male mortality rates from ages seventeen to thirty.

Outline why it may be inappropriate for the providers of death benefits to

change the assumptions they use for female mortality rates in the light of this
evidence. [5]

Page 68 © IFE: 2019 Examinations

 Exclusive use

45 Subject CA1 Paper 1 September 2017 Question 7 (part)

A professional orchestra that performs classical music concerts in a few

locations around one city is about to undertake its first international tour.
The objective of the tour is to raise the profile of the orchestra and generate
income both now and into the future.

Over three weeks there will be a twelve concert tour across six different
countries. There are about 150 people of many different nationalities in the
tour entourage including musicians and support staff.

The tour director has asked an actuary to model its future cashflows to
assess the tour’s profitability and/or likelihood of achieving its objectives.

(i) State the basic features of general actuarial models that are used to
project future cashflows.

You do not need to cover the steps or requirements for actually using or
building a model. [5]

(ii) Set out how sensitivity analysis would be used in conjunction with
modelling the cashflows relating to this tour, in order to assess the tour’s
profitability. [5]
[Total 10]

© IFE: 2019 Examinations Page 69

 Exclusive use

SOLUTIONS TO PAST EXAM QUESTIONS

1 Subject CA1 Paper 1 April 2008 Question 2 (part)

Items to consider – expense inflation assumption

 current rates of inflation, both for prices and earnings

 expected future rates of inflation
 the differential between the return on government fixed-interest
securities and on government index-linked securities, where such exist
 recent actual experience of the office or the industry as a whole
 consistency with other assumptions

2 Subject CA1 Paper 1 September 2008 Question 4 (part)

An asset-liability model (ALM) would be constructed.

All relevant cashflows that may arise should be included in the projection.
These will include both the individual’s assets (eg pension income,
investment income) and liabilities (eg day-to-day spending, one-offs).

It would be necessary to consider the level of matching on each projection.

If assets and liabilities are not matched, there is a risk that the individual will
run out of assets.

Allowance should be made for cash being put away in the form of savings.

Allowance should also be made for inflation.

The cashflows must be projected over an appropriate time horizon for the
cashflows. This is likely to be the expected life span of the individual or his
spouse (if longer than his expected lifespan).

The model should be run with a range of asset allocations. The amount of
liquidity required should be considered. It is likely that the current asset
allocation will change as a result of the redundancy payment.

Page 70 © IFE: 2019 Examinations

 Exclusive use

The sensitivity of the outcome to key factors should be taken into account, in
particular:
 life expectancy
 likelihood of ill health
 net returns on assets
 interest payments on debt
 lifestyle changes, eg unemployment
 variability of assets / liabilities.

Allowance should be made for risk mitigation techniques, eg life and health
insurance.

3 Subject CA1 Paper 1 September 2008 Question 5 (part)

(i) Factors to consider when analysing expense experience

Consideration should be given to whether the period under investigation was

typical and whether the experience is likely to be representative of future
experience.

For example, the period under investigation may have been affected by:
 abnormal events, such as a one-off fine to a regulator
 significant random fluctuations
 economic cycles and inflation
 trends in experience, eg due to automation of systems, which may or
may not continue into the future.

There may be heterogeneity within the past data, which could be different to
the heterogeneity that will exist in the future.

Different lines of business have different expenses, therefore it is important

to allow for:
 changes in the mix of business
 new products.

Overall, the quantity, and hence the credibility of the data may not be
acceptable.

© IFE: 2019 Examinations Page 71

 Exclusive use

Variations in the volume of business sold in the past and to be sold in the
future will affect the overhead expenses.

An adjustment will also need to be made for inflation.

Margins may be needed in future premiums to reflect uncertainty in the

expense experience, but not too large, as premiums need to be competitive.

(ii) Use of the previous expense figures

Advantages

Using the results of the previous analysis will be quicker and easier than
carrying out a new, full analysis.

This will also be cheaper.

It may be a suitable approach if there have been no significant changes to

the business over the period (and none are expected).

If expenses are only a small component of the overall premium, then this
method may be acceptable.

Whether the results are suitable or not depends on how long ago the
exercise was carried out – if it is recent then the results may be suitable.

Disadvantages

Past data is unlikely to be representative of what will happen in the future. In

particular, there may be changes in:
 volume of business sold
 mix of business sold
 new products, for which there is no past data to apply inflation to
 the services provided by the general insurer
 processes, eg greater automation may have led to lower costs
 sales channel and target market.

There may also be one-off expenses, such as a new marketing campaign,

that may not be included the previous analysis.

Page 72 © IFE: 2019 Examinations

 Exclusive use

Also, determining the right rate of inflation to use on different expenses may
be difficult, for example, salary inflation may be appropriate for some
expenses and price inflation for others.

A new analysis may be required for other purposes, eg setting provisions for
existing business.

Furthermore, legislation may require a full analysis of expenses for reserving

purposes.

If the results of the previous analysis are used and if they result in
inappropriate premium rates, then this will affect profits, eg through
uncompetitive premium rates, and the general insurance company may
make a loss.

4 Subject CA1 Paper 2 September 2008 Question 5 (part)

Problems with the quality or quantity of the data may lead to errors in the
assumptions used to calculate the risk premium.

The quality of the data may be poor due to problems with management
controls, verification processes, or data systems design.

Data from claims five or more years ago may be too old to be of relevance to
pricing business written now.

The term of the contract is twelve years, so the insurer will have no data
relating to the final two years.

5 Subject CA1 Paper 1 April 2009 Question 3 (part)

The company would have modelled its expected profits as follows:

 collect, group and modify data
 choose model points to represent the policies that are expected to be
sold (eg age, sex, premium size, choice of fund) based on similar
contracts if possible
 choose the form of the model, eg deterministic or stochastic
 stochastic modelling is most likely for unit-growth rate, particularly given
the guarantees given to the third fund
 set any relationships / correlations between variables, eg low investment
return may be connected to low inflation or high withdrawal rates
 determine the distribution functions of any stochastic variables

© IFE: 2019 Examinations Page 73

 Exclusive use

 choose parameter values for unit growth rate, non-unit investment

return, expenses (per policy and global), inflation, withdrawals, mortality,
new business levels
 construct a model of the unit fund and the non-unit fund. The expected
profits emerge from the non-unit fund as the charges less expenses less
cost of guarantees (on death or from the third fund) plus investment
return less the increase in reserves
 calculate the expected profit stream from each model point using both
best estimate assumptions and sensitivity testing
 multiply each model point by the business volumes expected each year
 calculate overall profitability from new and in-force policies each year
and deduct an allowance for overheads (eg design and launch costs).

6 Subject CA1 Paper 2 April 2009 Question 3

The three approaches to modelling are:

 buy a commercial modelling product
 reuse an existing model, possibly after modification
 develop a new model.

The factors to consider when choosing the approach to use are:

 accuracy required – a small one-off exercise may not justify a high
degree of accuracy, so modifying an existing model may be sufficient
 approximations – professional judgement should be applied to decide
whether a more approximate approach can be used for a one-off
 expertise available – if there is little expertise in-house a commercial
product may be more suitable
 number of times the model is to be used – this is a one-off, so the time
and cost required to build or buy a new model may not be justified
 desired flexibility – little flexibility is required for a one-off investigation, a
commercial product may be unnecessarily complex
 other uses – however, a more flexible model may be used if it can be
reused for other purposes or clients later
 cost – a small investigation may not justify the cost to build or buy a new
model
 data – the form of the model must be appropriate for the data available

Page 74 © IFE: 2019 Examinations

 Exclusive use

 purpose – using a commercial product or reusing an existing model may

not be able to do the job required
 uniqueness – if this investigation is unusual, then it may be necessary to
develop a new model.

7 Subject CA1 Paper 2 April 2009 Question 7 (part)

(i) Data

 current and retired member data (as at the valuation date)

– age
– date of entry / retirement (to determine service)
– salary / pension
– benefit entitlement
– marital status
 members who have transferred in / out of the scheme / died
 membership movements since the scheme was established
 market value of assets at both dates split by asset type
 market data (eg inflation and bond yields) for economic assumptions
 investment income
 contributions into the scheme
 benefits paid
 transfers of funds into or out of the scheme, eg members transferring
their benefits between schemes
 scheme accounts provide an audited source of data

(ii) Data checks

 reconcile membership against original data

 average ages / benefit levels should remain reasonable
 investigate unusual values, eg high ages
 spot checks on individual members
 check benefit payments against member numbers and movements
 check salary growth of members against payroll

© IFE: 2019 Examinations Page 75

 Exclusive use

 check contributions are consistent with salaries

 check investment income is consistent with asset values
 check asset holdings at the start and end of the period, adjusted for
sales and purchases

8 Subject CA1 Paper 1 September 2009 Question 4 (part)

(i) Reasons for mortality experience to be better than expected

 the insurer initially had no data for this sales channel, so may have
taken a particularly prudent approach to its pricing
 a high mortality assumption may have been chosen to keep volumes
low until the insurer had built up more experience
 errors in the pricing assumption due to a lack of data
 random fluctuations in experience, particularly if volumes were small
 underestimating trends in mortality improvements
 internet underwriting may have been more effective than expected
 the market targeted through the internet may have lower mortality
experience than expected

(ii) Effect of legislation on gender discrimination

Problems of heterogeneity

 without this legislation premium rates would be higher for males than
females
 the new single premium rate will be cheaper than the current rate for
males and more expensive for females
 the insurer is exposed to the risk that it insures a bigger proportion of
male lives than expected
 term assurance sales to males (females) may go up (down)

Pricing assumptions

 premium rates could be calculated using the weighted average of the

rates currently used for males and females
 the insurer could use weights based on its current mix of business
 but the mix of business is likely to change

Page 76 © IFE: 2019 Examinations

 Exclusive use

 the business mix will be highly uncertain at first

 so the insurer will assume a prudently high proportion of male lives
 the insurer should also consider the average mix in the industry as
quotes from different insurers can be easily compared on the internet
 assuming too many males will result in high premiums and low volumes
 assuming too many females will result in low premiums and losses
 the insurer may try to market its products to females, eg by placing
adverts in women’s magazines
 the insurer will monitor its experience frequently given the current high
levels of uncertainty

9 Subject CA1 Paper 2 September 2009 Question 7 (part)

(i) Model requirements

There is a similarity to Core Reading Question 1.

 the model should be fit for purpose  so should be capable of modelling

the cashflows resulting from these risks
 the cashflows modelled should encompass all significant features of the
business and the cashflow changes should the risk events occur
 for example, it should allow for the direct costs of building repair and
indirect costs such as reduced productivity
 the model should allow for the frequency and impact of each risk event
 regulatory requirements should be allowed for in the cashflows
 a long time horizon will give more accurate results
 the longer the time horizon the more costly the modelling exercise
 similarly, the frequency of cashflows modelled should be chosen so as
to achieve an appropriate balance between accuracy and cost
 the parameters / assumptions used should be realistic
 variables should be sensibly correlated
 the model should not be too complex and should avoid spurious
accuracy, but if it is too simple it may lack realism
 the model should allow for the amount and detail of data available
 there may be little relevant data if the risks occur infrequently

© IFE: 2019 Examinations Page 77

 Exclusive use

 a stochastic model may be useful as the risk events are highly uncertain
 a stochastic model would provide a distribution of results which could be
used to assess capital requirements
 scenario analysis would be useful as some of the risks (eg operational
risk) are hard to quantify
 sensitivity testing can check the sensitivity of the model’s results to
changes in parameter values
 the results of the model should be displayed clearly
 the output should be capable of independent verification, and should be
easily communicable
 the model should be capable of refinement
 the model should be well documented

(ii) Steps required to evaluate these risks using a model

 specify the purpose, ie the financial impact of these risks

 collect, group and modify data
 obtain advice from senior management and external experts on possible
risk events
 choose a density function for each variable modelled stochastically
 specify any correlations between variables
 set assumptions for each parameter
 it may be possible to use a pre-existing model, perhaps after
modification, or to purchase a commercially available model
 alternatively, construct a model from scratch, based on the expected
cashflows
 test goodness of fit against past experience
 attempt to fit a different model if the first model does not fit well
 run the model and perform sensitivity tests
 summarise the results
 compare results with the risk tolerance of the company
 if risk is unacceptable, model mitigation techniques until acceptable
results are produced

Page 78 © IFE: 2019 Examinations

 Exclusive use

10 Subject CA1 Paper 1 April 2010 Question 1

There is a similarity to Core Reading Question 1 and Past Exam Question 9.

Requirements of a model to assess the variability of future events

 fit for purpose

 adequately documented
 reflect the risk profile of future events and be capable of indicating the
variability of these events
 parameters must allow for all features that could significantly affect the
outcome
 inputs should take into account any special features of the future events
 workings of the model should be easy to understand and communicate
 results should be clearly displayed
 variables should exhibit sensible joint behaviour
 outputs should be independently verifiable and communicable
 model must not be overly complex or expensive to run
 capable of development and refinement
 different implementation methods should be available to assist testing
and parameterisation

11 Subject CA1 Paper 1 April 2010 Question 6 (part)

(i) Data required to calculate provisions

Policyholder details:
 age or date of birth
 sex
 rating factors, eg smoker status

Policy details:
 type of policy, eg endowment, annuity
 term to maturity
 duration in-force or start date

© IFE: 2019 Examinations Page 79

 Exclusive use

 sum assured
 bonuses declared
 premium details, eg amount, frequency, payment term

Financial data, eg inflation

(ii) Reasons to group data

 the model may be unable to handle a large number of individual policies

 it will be quicker, and hence cheaper, to run the model
 more deterministic scenarios or stochastic simulations can be run
 complex stochastic modelling may be required to value options
 increases in processing power reduce the need to group the data

(iii) Considerations to be taken into account when validating the

grouping of the data

Data groupings should reflect the profile of the full policy data so that the
calculated provisions are a close approximation to the true value.

Check to ensure that the key summary statistics are the same for the
grouped data and the individual data, eg number of policies, total sum
assured, total bonuses and total premiums.

The data needs to be summarised into sufficiently homogeneous groups.

If groups are heterogeneous then the grouped data will not give the same
results as ungrouped data in all circumstances, eg if assumptions are
changed.

Alternative groupings will need to be tried until the above checks are
satisfactory.

Page 80 © IFE: 2019 Examinations

 Exclusive use

12 Subject CA1 Paper 1 April 2011 Question 6 (part)

General

Some assumptions will be no different, eg investment return. The insurer

should focus on the ones that are most financially significant.

Claim frequency and amount

Motor insurance

Claim frequencies may be higher due to:

 higher rates of crime, eg theft, in student areas
 higher volumes of traffic in student areas (eg city centres)
 students being more careless.

Claim frequencies may be lower due to students driving fewer miles.

Claim amounts may be lower due to:

 students owning low value cars
 less severe accidents in student areas (eg city centres).

Buildings and contents insurance

The buildings cover may only extend to accidental damage since the
students are unlikely to own their own property.

Claim frequencies may be higher due to:

 higher rates of crime, eg theft – this may be exacerbated by students
living in shared accommodation
 students having a higher propensity to claim.

Claim amounts may be lower due to students having low value contents.

Their property may also include a higher concentration of certain items,

eg bicycles.

Students may have specialist items needing cover and the level of cover
and/or claims for these may be similar to the insurer’s existing market.

© IFE: 2019 Examinations Page 81

 Exclusive use

Travel insurance

University students may typically take different types of holiday, eg their

holidays may tend to be:
 cheaper, due to:
– cheaper destinations, transport and accommodation
– shorter durations
 longer, eg students might travel throughout their summer vacations or
go on gap years – such travel will require very different assumptions.

Claim amounts for cheaper holidays will typically be lower.

Claim frequencies and amounts may also be affected by the exact

destination and the risks they pose in terms of theft, illness etc.

Financial loss insurance

This may be significantly different to the insurer’s existing business as

different risks will be covered.

Health insurance / dental insurance / optical insurance / personal accident

Claim frequencies may be lower due to students being in younger and hence
in better health. (This may affect claim frequencies if the insurer sets
assumptions based on broad age bands.)

However, claim frequencies may be higher if they may take less care and
therefore be more prone to injuries / illnesses / accidents.

Claim amounts may be smaller as students would be unlikely to select the

highest level of cover.

Expenses

Overhead expense assumptions will depend on the volume of business sold

to the university students.

If claim frequencies are higher and claim amounts smaller for university
students, then claims handling expenses will make up a larger relative
portion of the premium.

Commission will depend on how the business is sold, eg the insurer might
set up branches in university campuses.

Page 82 © IFE: 2019 Examinations

 Exclusive use

Some expenses, eg underwriting costs, might depend on the level of sum

assured, which may be different for this business.

Lapse / renewal rates

Mid-year lapse rates for university students might be relatively low. This
might depend on whether it is single premium or regular premium business.

Renewal rates are likely to be low, eg as students graduate.

Business volumes

Business volumes will depend on how effectively students are targeted, as

well as the number of students that need insurance.

New business sales are likely to be highly seasonal.

Contingency margins

Margins might be higher due to greater uncertainty.

Profit loading

The profit loading will depend on factors such as:

 the competitiveness of the market
 whether certain products might be sold as loss leaders.

13 Subject CA1 Paper 2 April 2011 Question 2 (part)

New business volumes – these depend on:

 the number of births in the promotional period

 how well the offer is publicised.

Mortality rate – these depend on:

 the age profile of the prospective new mothers who apply for cover

 any increase in mortality rates for new mothers resulting from childbirth

 anti-selection from the prospective new mothers who apply.

© IFE: 2019 Examinations Page 83

 Exclusive use

14 Subject CA1 Paper 2 April 2011 Question 6 (part)

Process to establish a revised interest rate charging structure

An objective should be specified and should include:

 a quantifiable and measurable performance target, eg profit per credit
card customer, across all credit card business or across multiple lines of
business

 defined time horizons, which are likely to be short-term (1  2 years)

 quantified confidence levels for achieving the target.

The most appropriate charging structure will be the one that achieves the
highest profit, for a given time horizon and a given level of confidence.

The credit card company will need to specify a set of model points reflecting
both existing and expected new business.

Each group of policyholders will potentially be charged a different rate of

interest.

A trial interest rate structure will be selected for each model point, which will
reflect the average outstanding balance, repayment schedule and probability
of default – this may in turn depend on the interest rate itself.

Cashflows will be projected for each model point. This may be done:
 deterministically, which may be most appropriate for parameters that are
fairly stable and/or less financially significant

 stochastically, which may be appropriate for uncertain parameters.

A stochastic approach will yield a distribution of outcomes, however, it will be

impractical to model too many parameters stochastically.

The cashflows should be discounted back at a risk discount rate that reflects
the shareholders’ required rate of return and the statistical risk attaching to
the cashflows.

The resulting profit should be then be compared to the profit criteria.

Alternative interest rate structures should be investigated to see if they

provide a better outcome.

Page 84 © IFE: 2019 Examinations

 Exclusive use

Sensitivity testing should be carried out on the key parameters and the
choice of model points.

The interest rate structure chosen should also be compared to interest rates
offered by competitors.

15 Subject CA1 Paper 2 April 2011 Question 7 (part)

(i) Why ideal data may not be available

There is a similarity to Core Reading Question 64.

Ideal data may not be available for this purpose because:

 data may not have been captured at a sufficiently detailed level,
eg since the data was collected by an agent, complete data may not
have been gathered

 there may be insufficient data to provide a credible result – this may be

because the book is closed to new business.

(ii) How the company can use the limited data it has collected

There is a similarity to Core Reading Question 70.

Use summarised data, however, the reliability will be reduced because full
validation is impossible.

Even summarised data should be split into homogeneous groups; this must
be balanced against too little data in each cell.

For sparse data, group data into age bands – use sensitivity tests to help
choose exact age bands.

Exposed to risk should be grouped in the same way as deaths – use to

compare actual mortality to expected.

The company may be able to supplement its (limited) data with industry-wide
data, if it exists.

© IFE: 2019 Examinations Page 85

 Exclusive use

16 Subject CA1 Paper 1 September 2011 Question 6

(i) Criteria to assess if weather has disrupted or ruined the event

Consider the different types of weather that could disrupt or ruin the event.

For each type of weather assess how bad it needs to be before it disrupts or
ruins the event.

Calculate the probability that each weather type will occur on each date …

… and the probability it disrupts the event and / or that it ruins the event.

Assess the correlations between the weather events (eg snow, ice and very
low temperatures will be closely correlated).

Estimate the overall probability that the event would be disrupted or ruined
by any weather cause on a given day.

The organising committee can reject any date considered to have too high a
probability of being ruined by bad weather …

… or disrupted by bad weather (although a higher probability may be

accepted for disruption than ruin).

We might also reject dates that passed the above two tests if the combined
risk of the event being disrupted or ruined was high.

(ii) Appropriateness of available data for the prediction model

General points

The purpose of the model is to estimate the chances of various types of

weather in the town on particular dates in three years’ time …

… eg the model may predict that the probability of severe weather that
would ruin the event on 28 June is 10%.

A confidence interval could then be used to show the extent of any errors in
the modelling process …

… eg the 90% confidence interval for the probability of severe weather that
would ruin the event on 28 June may be 5% to 30%.

If the data is inappropriate for the model then the uncertainty in the results
will lead to a wide confidence interval.

Page 86 © IFE: 2019 Examinations

 Exclusive use

Relevance

The town is situated in a part of the country with particularly unpredictable

weather, so local data will be the most relevant.

The rest of the country may have very different weather conditions which
would make national data of limited use.

However, national data will more useful if the national weather exhibits a
strong correlation with the local weather.

Regional data may be useful if the weather within the region is fairly
homogeneous, eg if the region covered only the coastal part of the country.

Data extends over the last few centuries. However, the relevance of the
older data may be limited by trends such as global warming.

The most recent data is of most relevance to setting the starting point of the
projections.

The model then needs to be able to project this weather forwards over the
next three years.

Credibility

Weather records are available for hundreds of years, and so long-term

patterns in weather may be understood.

National and regional records may show patterns observed in local data.

National and regional data may be used to fill in any gaps in local records.

Fluctuations over time

The model will be used to look for weather patterns in the past that might be
repeated in the future.

The data may have shown weather cycles such as periods of hot dry
summers followed by periods of milder wetter summers.

The data may show trends eg increasing frequency of flooding and storms.

Predicting patterns requires more data than estimating the average level.
So the local data may be used to set the average and then national data
may be used to add trends and cycles to the average.

© IFE: 2019 Examinations Page 87

 Exclusive use

Data recording

There may be errors in the recording of the data.

It is likely that greater resources are available nationally than locally. The
national data is likely to be very accurate, with specialist equipment used,

The national data may also have been checked more rigorously.

Errors will also be more likely for the older records.

Errors may be introduced in the creation of a single database given that

records will have been kept in different formats over time.

Care is needed to convert these records to a consistent form, eg mixed use

of Centigrade and Fahrenheit.

There may be missing data, for example particular years or particular types
of weather may not have been recorded.

Local data will be more prone to omissions than the national data. It will be
a problem if recent years are missing as these are the most relevant.

It may be the case that the most extreme types of weather were not
recorded, but this is precisely the type of data that is of most use here.

(iii) Other considerations that could influence the choice of date

The actual date of the town’s founding would be a good choice, especially if
this date has been used for smaller scale annual celebrations.

It may be helpful if the event happens around the time of other local events,
giving visitors more choice and reason to visit the town.

Exact dates of other local events may need to be avoided so that resources
are available. For example availability of staff and equipment (eg marquees,
security) should be considered.

The dates of major national events elsewhere in the country should be

avoided if these would be a source of competition for visitors.

The dates of other major events with large TV audiences (eg World Cup
Final, Olympics) should be avoided if people tend to stay at home.

The timings of national and school holidays should be considered as these

will affect people’s willingness to travel to the celebrations.

Page 88 © IFE: 2019 Examinations

 Exclusive use

The date may need to be acceptable to the local council and the police, eg if
permission is required to close roads.

(iv) Difficulties in transferring financial risk to a third party

Difficulties in insuring the financial risk

The organisers could try to buy insurance to transfer the financial risk.

There may be difficulties in quantifying the risk, for example:

 estimating expected ticket sales for a one-off event
 determining what should be covered, eg should expected profits of local
retailers be covered?
 estimating the damage to the town’s reputation if the event fails.

It will be difficult to identify every possible source of risk to insure leading to

gaps in coverage.

It will be important but difficult to adequately define the risk events to avoid
disagreements between the organisers and their insurer.

For example, the insurer may argue that losses were incurred due to poor
marketing rather than bad weather.

Suitable insurance may not be available, as insurers may for example lack
data to price the risk and find it difficult to pool risk from similar events.

The cost of fully comprehensive insurance may be unaffordable, eg due to

the highly unpredictable local weather.

The organisers may not have the required expertise to arrange such a
complex form of insurance.

Difficulties with other risk transfer methods

Alternatively tickets could be sold in advance which would then provide a

known amount of income regardless of the weather on the day; …

… however, people may not be prepared to buy tickets on this basis, or may
demand a refund if the event is cancelled.

© IFE: 2019 Examinations Page 89

 Exclusive use

Finances could be raised through sponsorship and advertising …

… but it may be difficult to find sponsors to fund the event given the
uncertainty in the weather (and so the number of visitors their advertising
would reach).

17 Subject CA1 Paper 1 April 2012 Question 4 (part)

(i) Use of more complex models

Technical issues

Complex models should allow for a greater understanding of the risks

involved in financial services products.

This has led to greater to use of derivative products in order to control such
risks.

In turn, complex models are also appropriate for derivative products, which
can be complicated.

Complex models also allow financial services companies to tailor innovative

products to their customers, eg options and guarantees require complex
modelling.

Practical issues

Increased computer power means it is possible to build and run more

complex models, and to do so more cheaply.

More people have the knowledge and skills required to set up and maintain
complex models.

External issues

Products are more commonly offering attractive, but complex features such
as options and guarantees to increase marketability.

Even the assumptions required for modelling a simple financial instrument

(eg an ordinary share) are complex, so complex modelling will help to
explain the risks of these products better.

Regulatory requirements now require more complex models, eg for pricing

and reserving.

Page 90 © IFE: 2019 Examinations

 Exclusive use

(ii) Why complexity makes it difficult for the directors to understand

the risks

Directors are unlikely to be experts in derivatives / modelling.

The directors may find it difficult to understand the model and interpret the
results, and hence may rely on the design team to explain how the
products / models work, but they may use complex language and technical
terms.

The design team may find it difficult to explain complex concepts,

eg probability distributions, stochastic modelling etc.

The design team may only show the directors a summary of the full model,
given the significant volume of output produced. .

It should not be assumed that the design team have a perfect knowledge of
the company’s products and how they should be modelled.

Small errors in assumptions / correlations can have significant

consequences when using geared investments.

The design team may have an interest in the model being accepted, eg to
receive a bonus, which could lead them to understate the risks involved.

Hence, the directors may fail to understand the inputs, the model and/or the
results, specifically:
 which assumptions are the most important, and why

 the potential impact if implicit assumptions do not hold in practice, eg an

assumption of rational investor behaviour

 the directors may not understand how sensitive the results are to the
key inputs

The directors may be unable to ask the right questions of the design team
and instead may simply assume that the model is right.

Low likelihood but high impact risks may not be fully appreciated, which is a
significant issue for this insurance policy.

© IFE: 2019 Examinations Page 91

 Exclusive use

18 Subject CA1 Paper 1 April 2012 Question 5 (part)

Information is needed to establish:

 rating factors
 so as to divide the population into homogeneous subgroups
 to set assumptions in the pricing model.

We can then make similar assumptions for horses within each sub-group,
whilst ensuring consistent assumptions between sub-groups.

Demographic assumptions required include:

 Pre-retirement decrements – to understand the population size:
– pre-retirement mortality (before age 13)
– early retirement rates (before age 13)
– ill-health retirement rates (before age 13)  this is less important if a
cost-neutral basis is used
– new entrants – to understand how many new horses start racing at
each age

 Post-retirement decrements:
– post-retirement mortality – to understand the term of benefits
– late retirement – this is less important if a cost-neutral basis is used
 Other demographic information required:
– current age of each horse and the age at which they started racing
– gender of each horse
– category or class of races that each horse enters
– type of racing undertaken, eg jump races or flat races
– training centre / quality of racehorse (eg elite horses are more likely
to be retired for breeding).

Page 92 © IFE: 2019 Examinations

 Exclusive use

19 Subject CA1 Paper 1 April 2012 Question 7 (part)

(i) Determining the cost of mobile phone cover

The cost of claims will depend on the claim frequency and claim severity.

Claim frequency

Claim frequency will depend on:

 the proportion of gold-band holders with a mobile
 what proportion of these will lose their mobile phone or have it stolen or
damaged at the festival
 whether mobile phones are already covered by other insurance
 any claims controls in place
 type of festival and the presence of security or police.

Data to estimate claim frequency

Past data on the proportion of phones lost, stolen or damaged at this specific
festival in previous years would be very useful, but this may not be available.

Even if it is available, few people may have reported the loss of a phone in
previous years, since the insurance product wasn’t available.

Also, unless this is a very large festival, there may not be enough data for
the analysis to be credible.

Data from further back in time will be less relevant, since mobile phone
usage has increased significantly in recent years.

Alternative sources of data include:

 other insurers
 other festivals
 police records
 mobile phone retailers.

© IFE: 2019 Examinations Page 93

 Exclusive use

Claim severity

This depends on:

 the cover provided, eg:
– whether the benefit is a fixed amount / the replacement value of the
phone / cost of repairs
– other policy terms and conditions, eg exclusions, limits or excesses.
 moral hazard, eg visitors bringing more expensive phones to the festival
in the knowledge that their phones will be covered

 false claims, this risk can be reduced with claims controls.

Data to estimate claim severity

 repair or replacement costs  available from mobile phone retailers

 the type of phones owned by policyholders  available from home

contents insurance or mobile phone insurance policies (although this
may not be representative).

(ii) Claims experience for cancellation cover and mobile phone cover

Why experience is likely to be very different

Claim frequency is likely to be much more stable for mobile phone cover.

Claim severity for mobile phone cover may not be known (if it depends on
the particular phone).

Claim severity for the cancellation cover would be known (being the price of
the gold ticket).

Total claim costs:

 should be reasonably stable for mobile phone cover, as claim events
are relatively independent
 are likely to be very volatile for cancellation cover, eg one cancellation
will lead to refunds for many policyholders.

Why claims experience for cancellation cover is more difficult to predict

Claims experience depends largely on the weather, which is difficult to

predict very far in advance.

Page 94 © IFE: 2019 Examinations

 Exclusive use

There is the risk of anti-selection, ie if the weather looks bad just before the
start of the festival, visitors are more likely to buy gold tickets, rather than
silver.

There is also moral hazard, eg:

 the festival organiser may be more likely to cancel the event if the
visitors had the option to buy insurance

 the organisers may blame cancellation on ‘poor’ weather when the

underlying reason for cancellation is not weather-related.

There is subjectivity, as to what constitutes ‘poor’ weather.

20 Subject CA1 Paper 2 April 2012 Question 5 (part)

(ii) Reasons why past experience may not be a good guide to the
future

Data issues

The relevance of the company’s recent sickness experience must be

balanced against the need for sufficient data for an analysis to be statistically
credible.

Therefore the company must take sickness data from a long enough period
that it is credible, but not so long that it ceases to be relevant.

The company may not have sufficiently credible data given its size,
particularly when split by all the relevant factors (age, occupation, location,
smoker status etc).

Data may therefore have to be grouped.

The company may not have full, accurate sickness records, particularly, for
cases where no benefit was paid.

In the past, sick employees may still have come to work to avoid losing out
on pay.

Other data issues include:

 abnormal experience in the period to which the data relate, eg an
outbreak of swine flu

 random fluctuations, in particular given the size of the company.

© IFE: 2019 Examinations Page 95

 Exclusive use

Changes in environment

Changes in the environment that may mean that past sickness experience is
no longer relevant:
 an economic downturn
– people are more likely to get sick, eg due to stress
– people will be more worried about losing their job so will make an
extra effort to work even whilst sick

 medical advances, eg leading to better treatments / cures.

Changes in the class of lives

Past experience will be less relevant if there is a significant change in the

mix of people employed by the company:
 high staff turnover  past experience will tend to be less relevant; this
may be the case for unskilled factory jobs
 anti-selection  staff who believe they are more likely to get sick may be
more likely to join the company following the change in benefits
 a change in the mix of roles  eg a fall in the proportion of
manufacturing roles, leading to a drop in the number of injuries.

Claim inceptions and recoveries

Staff with minor illnesses (who may previously have still come to work) will
now be more likely to take sick leave. Therefore there may be many more
short-term absences for up to five days.

However, this might help to avoid contagious illnesses being passed around,
which might actually lead to reduced sickness rates.

The claims experience for the next three months of sickness may also be
higher, although this depends on how generous the discretionary benefits
typically were.

The incidence of absences will depend on the strictness of the doctors’

certifications compared to that of the immediate managers’ decisions.

Benefit payments could be significant for absences from work of up to three

or six months. However, since there were no formal benchmarks in the
discretionary system, there is unlikely to be any experience data at these
sickness durations.

Page 96 © IFE: 2019 Examinations

 Exclusive use

Sickness absences may be considerably shorter under the formal

arrangement, if the discretionary benefit was paid out for more than six
months.

There may also be a different mix of absences between manufacturing and

distribution roles, eg if the discretionary benefits were typically more
generous for (say) the distribution roles.

The decision to pay discretionary benefits had a subjective element, which

makes it less appropriate for statistical analysis.

There may be a significant difference between what different managers

deem to be ‘sick’.

Claim amounts

Claim amounts were discretionary prior to the introduction of formal sick pay
arrangements, whereas under the formal system, they are specified.

(iii) Difficulties with using other sources of experience data

Other sources of experience data include:

 national statistics
 industry data
 insurance companies offering sickness contracts.

Any other sources of data:

 are likely to be less detailed and less flexible than required
 are unlikely to be based on the same benefit structure – eg in terms of
level / term of benefits, so may differ in terms of sickness rates / benefit
amounts
 may not be based on exactly the same sickness definition
 may not be at the required level of detail, eg split by factory workers and
salespeople
 may not be based on experience in this geographic location /
socioeconomic mix of lives / type of industry
 may include different lives, eg self-employed individuals, rather than
employed staff
 may be out of date.

© IFE: 2019 Examinations Page 97

 Exclusive use

21 Subject CA1 Paper 1 September 2012 Question 3

(i) Usefulness of this data

The usefulness of the data will depend on the quality of the data and its
relevance, ie how representative it is of accident risks in the small country
concerned.

Quality

Good quality data should be complete, accurate and up-to-date.

The data needs to be sufficiently detailed which may not be the case if:
 there is insufficient volume of data
 the data is summarised.

Relevance

Data is needed to assess the potential frequency and size of claims.

The frequency and size of claims will be different here if, the frequency or
size of claims could increase if, for example:
 cars are different, eg poorly maintained or more expensive cars
 rules for driving are different, eg higher speed limits
 behaviours are different eg poorer standards of driving or higher crime
rates
 road types and conditions are different, eg poorer road quality.

(ii) Rating factors

The subsidiary could identify:

 proxy rating factors for age and gender
 other (true) rating factors that provide a sufficient assessment of risk.

Proxies for age and gender

Certain types of car might be preferred by women, men and/or certain age
groups.

Length of driving experience (eg number of years license held) might be

used as an indication of low age.

Page 98 © IFE: 2019 Examinations

 Exclusive use

Some postcodes might be largely populated by certain age groups,

eg young families or retired individuals.

Policy options / cover levels may be used as proxies if they are preferred by
men, women or certain age groups

However, all the above proxies will be imperfect. They may also be
challenged by regulators or consumer groups.

Other rating factors

Journey information might be gathered from tracking devices eg time of

day/night, speed, journey length, location, types of road.

However:
 past data will be limited or incomplete
 analysis may be difficult or inconclusive.

Other possible rating factors include:

 use of car (business or leisure?)
 annual mileage
 type of car
 number/cost of any accidents/claims
 occupation
 previous convictions.

The subsidiary could use an average rate based on assumed future

business volumes; …

… however this would open the company to anti-selection if other insurance

companies use a more sophisticated rating approach.

22 Subject CA1 Paper 1 April 2013 Question 2

See Core Reading Question 1 and Past Exam Questions 9 and 10.

© IFE: 2019 Examinations Page 99

 Exclusive use

23 Subject CA1 Paper 1 April 2013 Question 4 (part)

Need to project cashflows, eg:

 future premium
 investment income
 expenses
 claims:
– including claim inflation
– will depend on expected mortality rates / sickness rates / vet fees /
type of pet covered / treatment covered
– consider long courses of treatment separately, particularly if
renewals are guaranteed / premiums are capped
– any longer-term guarantees will make profits more sensitive to vet
fees
– allow for changes in treatments available.
 changes in reserves.

Allow for
 withdrawals on regular premium policies
 interdependencies between cashflows

 volume of business, eg will affect variable expenses and contribution to

overheads per policy

 mix of business, eg type of pet

 cost of capital.

The aggregated level of cashflows will form the expected results.

Consider sensitivity of results.

Page 100 © IFE: 2019 Examinations

 Exclusive use

24 Subject CA1 Paper 1 April 2013 Question 6 (part)

Internal data:
 from level premium term assurance business, to help set mortality /
persistency / expense assumptions

 for other pricing elements, eg profit contribution

 from other products, to help assess policyholders’ potential behaviour at

the option dates.

External data  to supplement the internal data and help price the reviewable
premium period:

 eg industry experience from other domestic insurers / reinsurers /

overseas markets
 but additional data is needed to understand differences in products /
underwriting / target markets / mortality.

Granular data will be required to understand the relationship between

mortality and persistency …

… this will depend on age, original term / premium, premiums after the level
term period and duration after the level term period.

25 Subject CA1 Paper 2 April 2013 Question 5 (part)

There is a similarity to Past Exam Question 11.

Claims data, to determine risk premium:

 claim frequency for each peril and claim amounts distribution
 data on any trends in claims experience
 needs to be obtained from an external source, eg reinsurer / industry
data.

Pricing data, to determine loadings:

 expenses, both direct and indirect, and expense inflation
 investment returns
 business volume / mix
 reinsurance premiums
 competitors’ premiums.

© IFE: 2019 Examinations Page 101

 Exclusive use

Policy data:
 exposure period, policy excesses / limits, destinations covered, number
of people covered, whether cover for hazardous activities is included

 rating factors, eg purpose of travel, age, health, gender (if permitted)

 distribution channel used (impacting on expenses incurred).

26 Subject CA1 Paper 2 April 2013 Question 6 (part)

(i) Uses of data

The main uses of data are:

 administration
 preparation of accounts and supervisory returns
 investment
 management information
 risk management
 product pricing / setting contributions
 setting provisions
 analysing experience
 marketing.

(ii) Checks on data

There is a similarity to Core Reading Question 63 and Past Exam

Question 7.

Check:
 asset data is complete, ie all assets are included in calculation, even
those purchased since the last valuation

 all assets still exist, eg bonds haven’t matured and are still owned by the
scheme

 no assets were double-counted

 market data agrees with investment income recorded in the accounts

 all assets have been valued correctly, eg correct market value /

discounted cashflow calculations

 no errors with units, for example in converting currency units

Page 102 © IFE: 2019 Examinations

 Exclusive use

 category of members is correct, ie deferreds / pensioners / movements

from active to deferred status
 membership data is complete, eg no missing members / dependants’
benefits are included

 salary-related increases have been allowed for, and comply with

company data

 benefits were paid as expected

 all planned events occurred, eg bulk transfers

 member benefits recorded correctly

 no impossible dates

 average values have changed as expected, eg average age

 membership data / movement data agrees with accounting data.

27 Subject CA1 Paper 1 September 2013 Question 4

(i) Scenarios for capital modelling

Scenarios may be derived from:

 historical experience

 judgement to determine hypothetical events

 the results of a stochastic model

 legislation.

(ii) Issues when building a capital model

Definition

A model is a cut-down, simplified version of reality that captures the

essential features of a problem and aids understanding.

Purpose

The purpose of the model is a key factor …

… because the model must be fit for the purpose for which it is being used.

The type of model required will depend on the purpose of the investigation.

© IFE: 2019 Examinations Page 103

 Exclusive use

For example:
 if there are financial guarantees to assess, then a more complex,
stochastic, model may be required …

 … whereas a model for checking premiums may be much simpler.

Cost

The expertise needed to build and run the model should be considered …

… alongside the funds available.

The model should be sufficiently accurate without being too complex.

Data

Consider the availability of data, eg for claims, premiums and expenses.

Check the data to ensure it is reliable and of sufficient volume to be credible.

Data may have been collected over various time periods and recorded in
different ways. The model should be compatible with the data.

Assumptions

Once the model has been decided and suitable data found then the various
parameters can be estimated.

Realism

The model needs to allow for all cashflows that may arise from premiums,
expenses, benefits and options …

… including those arising as a result of reserving requirements.

The model also needs to allow for the interaction of assets and liabilities.

The parameters must reflect the features of the company and the purpose of
the model, ie capital modelling …

… and the environment in which the company is operating.

Any model points used should represent the business adequately …

… so that the output from the model is reliable and credible.

Page 104 © IFE: 2019 Examinations

 Exclusive use

A good model will balance the need for reality and the need for simplicity.

Accuracy

The model should be rigorously tested and adequately documented …

… to minimise the risk of errors or unrepresentative features.

The results from the model should be independently verified …

… to give a further check on the accuracy and reliability of the advice given.

Communication

The model and results should be clearly communicated …

… otherwise it will be difficult for the end user to understand the results, and
there is a risk that inappropriate conclusions may be drawn.

Impact of poor modelling

Failure to comply with the above principles may mean that the model is
inappropriate …

… and leads to incorrect decision-making, eg pricing or reserving decisions.

28 Subject CA1 Paper 2 April 2014 Question 3

(i) Should industry lapse experience be used?

There may be good reasons why the company’s experience differs from the
industry, eg different customer base, sales channels and/or the data may
relate to different periods.

This would make the industry data will be less relevant.

Both data sets should be credible, as the company is large, and there should
be lots of industry data.

If there are concerns that the company’s internal data or the industry data
are wrong then this data should not be used.

For pricing, realistic assumptions should be used and a margin added for
uncertainty.

© IFE: 2019 Examinations Page 105

 Exclusive use

It seems sensible to use the company’s own experience rather than industry
data to determine realistic assumptions.

For reserving, prudent assumptions should be used which may be specific to

the company, or dictated by regulation.

Industry experience gives the higher lapse rate. Consider the implications of
this.

Type of contract

If the contract is regular premium, then an assumption of a higher lapse rate

than expected is likely to be prudent …

… as if policyholders lapse then initial expenses may not be covered and

losses made.

This is not true for single premium contracts.

Level of surrender benefit

Depending on the use of the results it may be prudent to assume higher or

lower lapse rates than are expected.

If the contract has a generous surrender benefit then more prudence (ie a
higher lapse rate) would be required.

Consider any guarantees on the surrender benefit.

Policyholders’ reasonable expectations may lead to guaranteed and/or

generous surrender benefits.

If there is a generous surrender benefit then a higher lapse rate assumption

may be more prudent.

If there is no surrender benefit, the company may make a profit on lapse …

… and so a lower lapse rate assumption would be more prudent.

Page 106 © IFE: 2019 Examinations

 Exclusive use

(ii) Validating the results of the internal experience analysis

Reconciliations

Reconcile the latest results with the previous results:

policies in force at the end of the period =
policies at the start of the period + new policies during the period
– exits during the year.

Reconciliations could be carried out for each product class and year.

Check the results of an analysis of surplus.

Cross checks

Check the data against other sources, eg check the accounting data against
the data used to administer the contracts.

Check for consistency between the claims data and exposed-to-risk data.

Reasonableness checks

Check whether the lapse rates are consistent with what was expected …

… and reasonable compared with previous lapse rates.

The drivers of lapse rate should be analysed. For example, consider how
lapse rates differ by sum assured, premium size, entry year, distribution
channel, calendar year …

… to explain the differences vs industry data.

Check whether there was an option to pay the premiums over a shorter
period, if so these policies may not have lapsed.

Ensure there is no ‘clustering’ of dates, with lots of policies ‘lapsing’ on the

same day suggesting a systems error.

Consider the lapse rate measure, eg the industry lapse data may be
unweighted whereas the lapse rates from the internal investigation may have
been weighted by premium size.

© IFE: 2019 Examinations Page 107

 Exclusive use

Check changes in the mix of business, eg if more policies are purchased by

older policyholders then this might lead to a reduction in the number of
lapses.

Spot checks

Spot check policies – determine when the policy lapsed, and check against
when the last premium was received.

Process

Check the data manipulation process …

… even if the data is incorrect, there may be an error in the output.

29 Subject CA1 Paper 2 September 2014 Question 3 (part)

Requirements of an actuarial model to place a value on the target

company

There is a similarity to Core Reading Question 1 and Past Exam Questions

9, 10 and 22.

Timescales and budget

The model should not be too complex, or take too long to build and run …

… or be unnecessarily expensive.

Realism

The model should be fit for purpose.

The model should reflect the risk profile of the business being targeted …

… with inputs appropriate to the target insurer …

… the discount rate should reflect the level of risk associated with the target.

The model should allow for all cashflows to/from the target business …

… including interactions between the assets and liabilities.

The model may need to be stochastic.

The model should allow for the effect of reserving requirements.

Page 108 © IFE: 2019 Examinations

 Exclusive use

The impact of acquiring the target on the insurer should be modelled.

Checking and documentation

The outputs should be verifiable.

The model should be adequately documented.

Communication

The output should be communicable to the management and shareholders.

Consider the sensitivity of results to the key features of the target …

… for example renewal rates and new business volumes.

Future acquisitions

It should be possible to reuse the model to value the next target.

30 Subject CA1 Paper 1 April 2015 Question 7 (part)

There is a similarity to Core Reading Question 1 and Past Exam Questions

9, 10, 22 and 29.

The model must be fit for the stated purpose.

The model must be adequately documented.

The model should reflect the risk profile of the annuity business.

The parameters must reflect the features of the annuity business.

The workings of the model should be easy to appreciate and communicate,

with clearly displayed results.

The model should have sensible joint behaviour of model variables,

eg interest rates and inflation.

The output of the model should be capable of independent verification.

The output should be communicable, for example to the regulator.

The model should not be too complex, such that the results are difficult to
understand.

© IFE: 2019 Examinations Page 109

 Exclusive use

The model should not be too costly or time-consuming to run.

The model should be capable of development and refinement, eg changing

regulatory requirements.

The model should allow for all the relevant cashflows that may occur,
including the bonus distribution policy of the mutual insurer.

Interactions between the cashflows, in particular assets and liabilities should

be allowed.

It may be appropriate to use a stochastic model, for example to model

investment returns or mortality improvements.

The model should be designed so that sensitivity testing can be carried out,
eg to changes in longevity or asset performance.

A decision needs to be made as to how often to calculate solvency during

the 20-year time horizon, eg half-yearly or yearly.

31 Subject CA1 Paper 2 April 2015 Question 3

Understanding / communication

The Chairman may find it difficult to understand the model and/or the
assumptions, and so may be reliant on the advice of others.

The Chairman should regard the model as a black box.

The analysts developing the model may have a vested interest in the
product.

The model is likely to be very complex, and so the Chairman may not fully
appreciate the model, significance of assumptions or be able to check the
results.

The analysts may be unable to communicate the assumptions, risks and/or

results of the model effectively to the Chairman …

… who may be persuaded without really understanding.

Model and assumptions

The model may not be perfect, ie only reflect key features in the interests of
simplicity.

Page 110 © IFE: 2019 Examinations

 Exclusive use

Many assumptions will have been made, some key assumptions may be
incorrect.

Assumptions will be needed about the efficiency of markets and the

rationality of investors.

Correlations between variables will need to have been specified.

It will be difficult to set some assumptions and determine appropriate

probability density functions.

An independent review of the model and assumptions is likely to be

appropriate.

Matching

The Chairman may not appreciate that there are some risks that are not
hedged.

It is unlikely to be possible to hedge all risks exactly, an approximate hedge

will be needed, investing in assets that do not precisely match the index.

It may be difficult to understand where the hedging strategies may fail.

The hedging strategies may also involve gearing, so assumptions about

interest rates are needed.

Risks that may not be hedged

The following risks will be difficult to hedge:

 interest rate risk
 currency risk
 default risk
 basis risk.

Tail risk

The Chairman may not appreciate the low likelihood but high impact risks
that could have disastrous consequences.

Such ‘tail’ risks are a major issue given that the aim is to achieve guaranteed
profits with no risk.

© IFE: 2019 Examinations Page 111

 Exclusive use

Pricing

The Chairman should also appreciate the need to cover expenses as well as
claims, and allow a margin for profit.

32 Subject CA1 Paper 2 April 2015 Question 7 (part)

Sources of data and their limitations

There is a similarity to Past Exam Question 24.

The main source of data is likely to be historical data from the first six
months since the airport opened.

There may not be sufficient data to be credible.

The data from the first six months may not be representative, eg airport staff
not fully trained yet.

Data from other airports could be used, but unlikely to be supplied by

competitors.

Reinsurers may provide data, and be able to offer expertise with pricing.

If the data is from a different source, the data may not be representative, for
example, the data may be based on different:
 luggage / flight volumes
 locations
 baggage handling systems
 geographical locations / cultures.

Many of the general limitations of data apply, for example:

 abnormal or random fluctuations
 changes in experience over time
 heterogeneity in the data
 changes in the balance of the homogeneous groups.

It is important to balance the amount of data gathered vs its relevance.

Page 112 © IFE: 2019 Examinations

 Exclusive use

33 Subject CA1 Paper 1 September 2015 Question 5 (part)

(i) The objectives of and requirements for building a model

There is a similarity to Core Reading Question 1 and Past Exam Questions

9, 10, 22, 29 and 30.

(ii) Decisions to be made when determining the modelling approach

Cashflows

The model needs to allow for all cashflows that may arise in future, so Helen
needs to decide which sporting events to model.

The model should allow for the cashflow arising from Helen’s lifestyle, and
Helen needs to decide whether some or all of the cashflows are to be
included.

Helen needs to consider the interactions between the different cashflows,

and whether and how to allow for them.

Data

Helen needs to decide what data to use and how many years’ data to use.

Showing the uncertainty of results

Helen needs to decide whether to use a stochastic or deterministic model.

The model should be able to evaluate the probabilities of wins and losses …

… and so stochastic techniques may be appropriate as they allow for

probabilities …

… which gives Helen a guide as to the likely distribution of the capital

requirements.

Helen needs to decide the extent to which sensitivity analysis should be

used to analyse the variability of the results of the model to the assumptions
chosen.

Model output should be in an appropriate format that Helen can easily

understand.

© IFE: 2019 Examinations Page 113

 Exclusive use

Assumptions

Helen needs to decide what assumptions should be used.

Model

Helen needs to decide the number of simulations (if a stochastic approach is

used), the choice of parameters and how parameters will be estimated.

Helen will want the model to strike a balance between realism (with the risk
of increased complexity) and simplicity (so easier to build, understand and
check) …

… so Helen needs to decide on the importance of accuracy and the budget

available.

Helen needs to decide what projection period should be used …

… considering the time taken to run the model (a longer projection period
means longer run times) …

… and the required accuracy of the results (calculating the cashflows more
frequently should be more accurate but more complex).

When deciding which approach to take, Helen should consider the expertise
and resources she has access to …

… and her budget and the cost of the various models.

The model is likely to be unique to Helen’s personal circumstances, so she

may prefer to develop her own model …

… which may take a relatively broad-brush approach using judgement.

Source of the model

Helen needs to decide how she will obtain the model. She could:
 buy a commercial modelling product
 modify and use an existing model
 develop a new model from scratch.

Page 114 © IFE: 2019 Examinations

 Exclusive use

34 Subject CA1 Paper 1 September 2015 Question 7

(i) Uses of data by a financial institution

There is a similarity to Past Exam Question 26(i).

(ii) Data requirements for setting premiums for funeral cost plans

Claim timing

Every policy will pay out if the contract is still in force after ten years (as
every individual will eventually die) …

… so the key issue is when the payment will be made …

… if there is data of sufficient quality and quantity, this can be based on the
experience of existing policyholders.

The funeral director could compare its mortality rates with the general
population …

… noting that there may be significant anti-selection …

… as, for example, the policy is likely to attract those in poor health.

The funeral director is likely to want data to analyse mortality by:

 gender (especially if regulation requires unisex pricing)
 age
 duration (to understand how many years’ premiums are paid before the
benefits are payable).

The funeral director should also consider data available on mortality

improvement rates.

Claim amount / funeral costs

A key factor in setting the premium will be the claim payout.

The amount to be paid will depend on the cost of services provided by the
funeral director …

… which may be the same or similar for all, if services are standardised.

© IFE: 2019 Examinations Page 115

 Exclusive use

Data will be needed on the costs of different types or levels of funeral, if the
product does not provide a standard benefit.

Funeral cost inflation

Data will be needed on historical rates of increase in funeral costs (for

example relative to price inflation).

Volumes

Data will be needed about the population of the city …

… and the proportion of those individuals who are likely to take up the
product.

The take-up rate is likely to vary for different groups of people …

… depending, for example, on their age, gender wealth, location, occupation

and health status.

Certain groups will be less interested in purchasing a funeral plan, for

example:
 younger people may feel there is no need to save for a funeral yet
 those on very low incomes may not be able to afford to save for a
funeral
 wealthy people may not feel that a funeral plan is necessary.

Cancellation rates

Data will be needed about lapse rates …

… at different ages and durations …

… as early lapses could result in a loss if initial expenses are not covered …

… and later lapses (before 10 years) should generate a profit.

Expenses

Data will also be required on the expenses of managing the funeral

product …

… and on expense inflation.

Page 116 © IFE: 2019 Examinations

 Exclusive use

Expenses may be split into fixed overheads and per policy expenses.

Investment return

Data will be needed about historical investment returns, as the premiums will
need to be invested for some time.

Competition

Data is needed on competitors’ premiums, to ensure the product is priced

appropriately.

(iii) Data requirements of an insurer offering whole life assurance

policies nationally vs the funeral cost director

Similarities

Both products are assurances paying out a lump sum on death of the
individual …

… and so the key data will be similar.

The key data required for both relates to claim timing, amounts, volumes and
expenses.

Differences

Claim timing

The insurer has national coverage, so data on the location of individuals will
help with differentiating pricing by area …

… this is not as relevant for the funeral director as all individuals are from the
same city (ignoring variations within the city).

The funeral director is more likely than the insurer to suffer concentration
risk, with exposure to just one city …

… and so should consider whether the data has been distorted by

catastrophes.

As the insurer sells nationally, this may expose the insurance company to
greater levels of selection unless it varies its prices by region …

… and so it will need more data to determine homogenous groups.

© IFE: 2019 Examinations Page 117

 Exclusive use

Collect data on the distribution channel to identify whether there is the risk of
anti-selection by sales channel (whereas all funeral cost plans are sold
directly).

The insurer is more likely to carry out underwriting, and so may need more
data to support its underwriting process.

Claim amount / funeral costs

A key difference is the size of the lump sum benefit under the two
contracts …

… particularly if the funeral director offers a standard service, as the sum

assured under whole life assurances could be for a wider range of values …

… so the insurance company will require additional data relating to the sums
assured.

Claim inflation

Unlike for the funeral director, there is no need for the insurance company to
consider inflation of benefits (unless the sum assured is index-linked).

Volumes

Due to a potential larger volume of products the law of large numbers is

more applicable to the insurance company …

… this may reduce the volatility of experience, and so the data collected may
enable assumptions to be set more precisely.

Cancellation rates

The funeral director and insurer will have different requirements relating to
lapse data.

Lapse rates are likely to be more sensitive to duration in force for the funeral
director, so data on lapse rates at different durations is likely to be more
useful.

Expenses

Given the greater volume of business sold, the insurer is also likely to have
more data on expenses, which should be more predictable.

Page 118 © IFE: 2019 Examinations

 Exclusive use

Expenses may differ by distribution channel for the assurance product, so

the insurer may need data on how they differ (whereas all funeral costs
plans are sold directly).

The insurer may also need to data to analyse contributions to overhead

expenses.

35 Subject CA1 Paper 2 September 2015 Question 6 (part)

(i) Assumptions to value the defined benefit pension scheme

Financial assumptions
 Investment returns

 Salary growth

 Price inflation (if the increases to benefits in payment or deferment are

linked to inflation)

Demographic assumptions
 Pre-retirement mortality (members and dependants)

 Post-retirement mortality (members and dependants)

 Mortality of ill-health retirees

 Mortality improvement rates

 Withdrawal rates in good health

 Retirement rates due to ill health

 Marital statistics

(ii) Suitability of historical data to determine each assumption

Investment returns

Historical data will be useful relating to the returns on assets.

Historical data will be available on the dividend yield on equities.

However, investment returns can be volatile over time, need to consider how
typical the past data is of the expected future. It may be that only the most
recent data is relevant.

© IFE: 2019 Examinations Page 119

 Exclusive use

Earlier data can be useful to identify the impact of particular economic or

fiscal events …

… this may require a long period of data to be assessed.

The pension scheme is likely to have long-term liabilities, so it is important to

look at data over a sufficiently long period.

Salary growth

Historical data relating to salary growth can be useful from:

 the company or the industry, or
 the country.

Adjustments may need to be made to the historical data if future salary

increases are expected to be different to the past …

… should request the sponsor’s views on expected future salary increases.

Data from the industry or country may not be directly relevant to the
members of this scheme …

… therefore adjustments may need to be made to the historical data …

… to ensure it reflects the characteristics of the membership.

Price inflation

Inflation tends to fluctuate significantly over time.

Historical inflation data is unlikely to be very helpful for setting an

assumption for the future …

… current levels of the index are more likely to form a good guide.

Indications of future inflation can be determined from:

 government policy and projections

 the gap between the yield on fixed-interest and index-linked bonds of an

appropriate (long) term.

When choosing data, need to reflect the appropriate measure of price

inflation relevant to the scheme, eg retail price inflation or consumer price
inflation.

Page 120 © IFE: 2019 Examinations

 Exclusive use

Mortality

Historical data is likely to be available from recent investigations by the

scheme and:
 insurance companies or industry data

 national statistics.

Data from these two external sources is unlikely to be directly relevant to the
scheme’s members …

… adjustments will need to be made to the data to reflect such differences.

Significant adjustments may need to be made to other sources of data if the

scheme membership is particularly non standard.

Medical advances are likely to mean that mortality will improve over time,
and this will not be reflected in historical data.

Historical data can be used to project future mortality improvements.

Significant emphasis is likely to be placed on the most recent data …

… trends in the past data and their underlying causes will help determine the
allowance for future improvements.

As ill-health and early retirement are unlikely to be very common, there will
probably be a lack of data available.

Withdrawal rates in good health

Historical scheme experience may be used …

… but it may be volatile (eg affected by a past redundancy exercise) or out

of date.

The company can provide information on its likely future withdrawal rates …

… some information however may be confidential, eg future redundancy

plans.

© IFE: 2019 Examinations Page 121

 Exclusive use

Retirement rates due to ill health

Ill-health retirement rates will be affected by changes to the definition of or

benefits available on ill health or changes within the business that may affect
the likelihood of suffering ill health …

… the company can help provide this information.

Marital statistics

The importance of this assumption depends upon how marital status affects
the benefits paid, eg whether marital status at retirement or death or leaving
is important.

Historical data may not be relevant due to the impact of common-law

marriages and civil partnerships.

36 Subject CA1 Paper 1 April 2016 Question 5 (part)

The importance of mortality for term assurance products

Term assurance pays out only on death within the term …

… so actual experience relative to the mortality rate assumption is critical in

determining the profit made.

If too low a mortality rate assumption is used when pricing / reserving, then it
is likely the insurer will make a loss.

Significant profits may be made if there are fewer deaths than assumed …

… however, there are restrictions on assuming a high mortality rate …

… since it means the product may not be competitive.

The mortgages may be repayment mortgages and therefore the term

assurance plans may be decreasing …

… therefore early deaths would be the main concern.

The contracts are regular premium, so on early death it is unlikely sufficient

premiums have been paid to recoup initial costs.

Page 122 © IFE: 2019 Examinations

 Exclusive use

Profits are very sensitive to mortality experience, since …

… the sum assured may be very large …

… the reserves held to back each contract are small …

… and unexpected deaths have a material impact on profit.

The impact of unexpected additional deaths is particularly marked where

policyholders are young, and so expected death rates (and therefore
premiums and reserves) are low.

The business may be subject to concentrations of risk,

eg geographically, this can increase the volatility of profits.

The contracts are likely to be without-profit, so any poor experience lies

solely with the insurer.

The mortgages and hence term assurance contracts are likely to be long
term, and there will be uncertainty about mortality in the long term.

37 Subject CA1 Paper 1 September 2016 Question 5

(i) Data limitations that determine whether the results should be used

Social conditions may have changed

Consider whether the past data is likely to be representative of expected

future mortality experience.

For example, consider whether future policyholders are expected to have the
same characteristics as existing policyholders …

… it may be that future policyholders will be less likely to smoke.

Trends over time

Three years may be too short a time to consider mortality experience, where
trends tend to be much longer term.

Need to check how the trend has evolved over the period, and whether it is
expected to continue …

… eg has the trend been uniform over the three years or was it greater for
certain periods?

© IFE: 2019 Examinations Page 123

 Exclusive use

This trend in smoking-related illnesses may not affect all types of

policyholder in the same way, eg it may vary by socio-economic groups,
gender, age etc …

… and the data relating to those groups may be insufficient to be credible.

Need to check whether this trend is specific to the annuity business written
or is it a more general population effect; …

… this will help determine whether the experience is just an abnormal

fluctuation or a real trend.

There are likely to be medical advances, reducing the impact of smoking on

health.

Errors in data / lack of data

Past data may contain errors and be of a poor quality.

The type of data may not be appropriate, eg mortality data split by number of
lives rather than amount of annuity.

Check whether the results are within statistical tolerances in terms of

business written and the experience of the general population.

There might not be sufficient appropriate data for it to be credible if, for
example, the insurer is small, annuities are not the main product sold or
annuities have not been sold for long.

Recording differences / lack of detail

Data may not have been collected at a detailed enough level, eg annuities
are not usually underwritten at outset, so there may be no information on
smoker status at that stage.

Information on smoker status may only have come to light on death.

Data on smoker status may lack detail, eg the number of cigarettes smoked
per week.

Heterogeneity with the group

There may be changes in the market for impaired life annuities.

Page 124 © IFE: 2019 Examinations

 Exclusive use

Consider whether the insurer has sold / will sell impaired life annuities.

If the insurer has changed to selling impaired life annuities during the period
of the investigation then the assumptions may already have been altered.

If information on smoker status was collected when the annuities were sold,
then it could be used to determine how many smokers are left of the annuity
business sold …

… ie may now have a much healthier population (of mostly non-smokers)

following the deaths of (mostly) smokers.

(ii) Data sources to set mortality assumptions for large annuities

Tables

Standard mortality tables may be produced.

Data may be available relating to defined benefit pension schemes.

Reinsurers

Reinsurer data may be available, whether it is a suitable source depends

upon the credibility and relevance of the data.

Data could be sourced from several reinsurers to check for consistency in

the results.

Industry data

There may be published industry tables, although need to check for

differences in target market and the data may not be sufficiently detailed.

The definition of ‘large’ benefits may be different across insurers.

National statistics

National statistics may be produced giving details of mortality experience

split by different socio-economic groups or location.

Experience investigations on existing contract

If the insurer has written a substantial amount of annuity business then it

may have sufficient past data on these large annuities that make up 20% of
the portfolio.

© IFE: 2019 Examinations Page 125

 Exclusive use

However, it is likely that the insurer will not have sufficient data to base the
investigation solely on this source.

Whether experience on existing contracts is relevant depends on any

changes to the annuity market, eg increased availability of impaired annuity
contracts.

Whether the data is credible depends on how many policies make up the
20% of the portfolio.

Consider whether the definition of ‘large’ has changed over time.

Similar contracts

The insurer may have sold large policies for other contract types, although
this data may not be particularly relevant.

Abroad (data from overseas contracts)

Overseas data may be available, but may lack relevance as the population
will be exposed to different social and economic effects.

(iii) Whether the mortality assumption for large annuities should differ

Differences in mortality experience on larger policies

It is likely that policyholders with large annuities live longer, eg they are likely
to be able to afford better healthcare and food.

The insurer will need sufficient credible data to determine how material the
difference is.

Need to consider extent to which different sized policies are affected, a

scaled assumption may be needed, ie greater adjustment the larger the size
of the annuity.

Likely future experience

Need to determine both the base table and future improvements to

understand whether the current difference in experience is likely to be
repeated in the future …

… as, for example, medical advances and better nutrition may be

increasingly available to other groups in the population, narrowing the gap.

Page 126 © IFE: 2019 Examinations

 Exclusive use

Impact of changing the assumption

Mortality is a key assumption for pricing annuity business, as it is a material

assumption it is appropriate to adjust it.

If the mortality assumption is made lighter for larger annuities, then it is likely
the assumption for standard annuities should be made heavier …

… to maintain the current average.

A lighter mortality assumption would increase the premium and likely reduce
business volumes for larger annuities; …

… need to consider whether the business is already profitable on the current

(lower) premium.

The premium rates charged by competitors should be considered, if

competitors do not use a lighter assumption then the insurer may be priced
out of the market.

Some policyholders with large funds may have brought several smaller
annuities, rather than one large one, the insurer needs to identify this from
the data.

38 Subject CA1 Paper 2 September 2016 Question 2 (part)

Technical issues

Modelling

Need to build a model to understand the expected future cost of each option.

This is done by comparing the value of the benefits before and after the
option is exercised.

Option pricing techniques could be used.

Assumption setting

Investment returns and longevity are the key assumptions.

In setting the assumptions, the scheme will need to consider the strength of
basis; including a margin for prudence will reduce the risk of the scheme
suffering future funding difficulties.

Consider the scheme's valuation basis.

© IFE: 2019 Examinations Page 127

 Exclusive use

Perform sensitivity testing.

Practical issues

Selection risk

Consider the risk of anti-selection, eg by individuals who are in poor health.

Categorise members into different groups and offer different factors to

members in different groups to prevent anti-selection.

An individual in poor health is likely to take up both the pension increase

option and the option to commute members’ pension for dependants’
pension, leading to a concentration of risk.

Data

Understand the likely take-up rate of the options, by gathering relevant data.

Gathering data about the health of the member and dependant will help the
scheme price the option appropriately.

Obtain full data, eg the age of the spouse is needed if additional dependant’s
pension is being purchased.

Administration

Consider the administration costs of calculating the option terms, an

appropriate allowance should be made for these costs.

External issues

Competition

Consider the factors used by competitors’ schemes.

Regulation

Consider any regulatory restrictions that may apply, eg any statutory

minimum pension increases in payment.

Risk management

Consider the impact that any risk management would have on the
assumptions used, eg if underwriting has been done to determine whether
members are in poor health.

Page 128 © IFE: 2019 Examinations

 Exclusive use

39 Subject CA1 Paper 2 September 2016 Question 7 (part)

(i) Advantages and disadvantages of ELL

Advantages of using the ELL system vs spreadsheets

Model

ELL may be less prone to model errors than spreadsheets …

… as it is already extensively used across the industry.

ELL should have been thoroughly tested.

There is better control over the model with ELL – as individual users can’t
amend ELL code but can amend the spreadsheets.

Changes to ELL should also be better controlled and tested.

ELL will have been externally audited, subject to detailed peer review and
more formal governance controls.

There is less chance of key-person dependency that can arise with

spreadsheets and less chance of ‘version control’ issues.

Stochastic modelling

Stochastic modelling, where required, should be more accurate and quicker

with ELL.

It may be easier to cover a full range of scenarios using ELL …

… and to allow for interactions and aggregate product lines together.

Data

ELL may be less prone to data errors than spreadsheets, as the data is fed
in automatically …

… which should also mean it is quicker to input the data into ELL.

There may be more confidence in the completeness of the data using ELL.

Documentation

ELL may be better documented.

© IFE: 2019 Examinations Page 129

 Exclusive use

Output

It may be easier / cheaper to independently verify the outputs from ELL.

Workings of the model may be easier to appreciate and communicate.

The outputs from ELL are likely to be clearer, as more standardised.

It may be less likely that there are material errors in the valuation results if
ELL is used.

It should be easier with ELL to achieve consistency across products and

valuation dates, so results should be quicker and easier to interpret and
communicate …

… and quicker to aggregate and disaggregate.

It may be easier with ELL to change the frequency of output.

Purpose

A model that can handle the entire business is more consistent with
economic capital methodology …

… and easier to use as part of enterprise risk management.

It may be easier to use the model to help with pricing.

Regulation

The regulator may have more confidence in results from ELL.

If ELL is well-understood by the regulator complying with professional

guidance on modelling or data then it should be easier and cheaper to
comply with regulation.

Less intervention may be required and less capital may need to be held.

The directors may also have more confidence in being able to sign off the
results.

Supervisory valuations may require such individual policy valuation.

Page 130 © IFE: 2019 Examinations

 Exclusive use

Practicalities

Given ELL is already being used for most products, it seems most efficient to
use it for all of them.

More support may be available, eg the consultancy may offer a helpdesk.

Changes can be rolled out once to all products under ELL, whereas each
spreadsheet would need to be changed separately.

ELL may offer a more user-friendly interface.

Speed

It could be quicker to run ELL …

… as large numbers of policies are impractical to manage using

spreadsheets.

Disadvantages of using the ELL system vs spreadsheets

Cost

ELL is likely to be more expensive.

There may be ongoing license fees to pay for using the software.

The hardware necessary to run the system will entail further costs.

Flexibility

As ELL is externally produced, it may be less flexible.

It may take longer to develop and refine ELL and require specialists to code.

It may not be possible to make quick ad-hoc / experimental changes to the

ELL model.

It may take longer to get additional runs / look at alternative scenarios,

eg for analysis of surplus …

… as it may be necessary to wait for ELL processing time to become

available, or to go through the central team …

… especially if the software can be used only by a limited number of people.

© IFE: 2019 Examinations Page 131

 Exclusive use

It may take a long time to get approval before any developments can be
made.

The insurer has no control over upgrades to software.

The support offered by the actuarial consultancy may be poor.

Control

The company is exposed to the risk of failure of the software / consultancy.

Transparency

As ELL was built externally, the modelling process may be less transparent,
as cannot see the model code or data and so may be less likely to spot
problems.

ELL may be less well understood than a spreadsheet, making debugging

more difficult.

ELL may introduce systematic errors affecting multiple processes.

ELL may be treated as a ‘black box’ and give the insurer a false sense of
security.

(ii) Steps and investigations before providing written approval

Project

The actuary should set key milestones at the start of the project.

Progress should be monitored regularly during implementation to ensure

completed in time for the next valuation.

Critical-path analysis can be carried out to ensure key dependencies are

being managed.

Time should be made available for peer review.

The actuary should check that all necessary valuations can be produced
from ELL for the next valuation …

… and that the data team can provide the required data in the required
format to the modelling team.

Page 132 © IFE: 2019 Examinations

 Exclusive use

The actuary should check that contingency plans are in place in case ELL is
not ready in time.

The valuation actuary needs to be happy with the results from ELL …

… although implementation of ELL may proceed, subject to workarounds …

… in which case the necessary workarounds must be in place for the next
valuation.

Data

The actuary will want to be satisfied that data used for the ELL valuation is
correct and complete and consistent, ie it can be reconciled to the data used
in the spreadsheet.

To ensure consistency, a parallel run of both data systems at the same date
should be carried out and ideally the data will match exactly.

Model

The actuary will want to be satisfied that the calculation method is accurate
on ELL and that the valuation result can be reconciled between ELL and
spreadsheet.

Both systems should be run at the same date, with the same data and
assumptions and the valuation result should be the same.

Any material difference needs to be investigated, documented and

understood or resolved.

The team involved in coding should document their formulae.

The valuation actuary should compare the coding with the spreadsheet …

… this may not be an independent check, if the coding came from the same
source.

It would be useful to spot check ELL and spreadsheet results for some
sample policies.

Stress tests and sensitivity tests should also be conducted.

Data

If possible, data for each individual policyholder should be cross-checked.

© IFE: 2019 Examinations Page 133

 Exclusive use

If each individual data point is not to be reconciled, then summary statistics

could be compared.

The data sets may not reconcile, in which case further investigations are
required to work out which data set, if either, is accurate.

Any material differences between data sets should be investigated,

documented and understood or resolved.

Output

It will be necessary to ensure the valuation result is correct.

The actuary should review the output from ELL to ensure it contains all the
required information …

… and can be downloaded, summarised or manipulated easily.

It should be checked that a reserve was calculated for all policies input into
ELL.

It should be demonstrated that the valuation result remains unchanged for

the lines of business previously run on ELL.

The valuation actuary should also review the run-time statistics to ensure the
ELL can deliver results quickly enough.

40 Subject CA1 Paper 1 April 2017 Question 4 (part)

Projection of the expected profits

Model

A model will be needed to project the insurer’s expected future experience.

The results of the initial pricing models can be combined to build a complete
model of the provider’s future revenue accounts.

A model point approach may be adopted, projecting profits for representative

policies and then grossing up the projected results by the number of policies
expected to be sold in each future year.

Assumptions

Assumptions will be needed for each revenue item, eg new business levels,
claim rates and expense inflation.

Page 134 © IFE: 2019 Examinations

 Exclusive use

The assumptions used for the projection may be those used when the
products were developed.

The basis used needs to be consistent, eg there needs to be a sensible

relationship between investment returns and inflation.

Cashflows

The cashflows in the revenue account will be projected forward.

The cashflows include premiums plus investment income less claims less
expenses.

The projection needs to allow for the impact of outstanding claims on

existing business.

Liability business can be long-tail, so there may be claims yet to be reported

and settled.

The projection also needs to allow for the number of existing policies
expected to renew each year.

41 Subject CA1 Paper 1 April 2017 Question 6 (part)

(i) Practical problems arising from a lack of data

The lack of data makes it difficult for the insurer to determine appropriate
prices and set aside appropriate reserves and capital.

It will be difficult for the insurer to set assumptions relating to:

 mortality rates
 expense loadings
 business volumes
 business mix.

External sources of data will be needed, eg industry statistics or reinsurer

data.

External sources of data may lack credibility and relevance, and may only be
available at a cost.

The insurer may know what the key rating factors are but will lack the data to
understand how they affect the particular target market.

© IFE: 2019 Examinations Page 135

 Exclusive use

A lack of data may cause difficulties in modelling and designing systems.

(ii) Commercial implications of data problems and policy features

Commercial implications of poor data

Price too high

The uncertainty in the data may lead to large contingency margins when
pricing.

Hence premiums may be high relative to competitors, so that business

volumes are low and fixed costs are not covered.

High business volumes are likely to be needed to cover overheads on this

product.

Additional costs may be incurred in accessing external data, and these will
be reflected in the premium.

However, the target market may not appreciate that the product is
expensive, so still purchase it.

Price too low

Poor data may lead to the product being under-priced, attracting large
volumes of potentially loss-making business.

The institution may be prepared to accept losses in the short term, as it

‘buys’ market share to increase its presence in the market and gather data.

Incorrect pricing structure

Poor data may lead to an inappropriate pricing structure, with good risks
charged too much and poor risks charged too little. This can lead to losses.

A lack of data and expertise may mean the institution lacks an

understanding of the current state of the market and extent to which there is
scope to gain market share.

Commercial implications of policy features

Minimal underwriting

A very limited level of underwriting can lead to selection risk.

Page 136 © IFE: 2019 Examinations

 Exclusive use

To keep expenses low, minimal claims controls may be in place, leading to

claims being greater than expected.

Low cost product and the target market

The idea of the product being low cost may conflict with the anticipated high
mortality of the target market.

If a high premium is charged to reflect the higher risk, then the benefit on
death may seem poor value for lower-risk customers, eg it may be less than
the regular premiums paid for those who live for a long time.

This may lead to reputational risk or possibly regulatory intervention.

Regulatory intervention may be more likely, given the target market may be
viewed as financially unaware and vulnerable.

Fixed benefit

This is a whole of life policy, with potentially a long term, so the fixed benefit
may be eroded by inflation and be insufficient to meet funeral expenses.

42 Subject CA1 Paper 2 April 2017 Question 4

(i) Assumptions to value a pension scheme

 pre-retirement mortality
 post-retirement mortality
 future improvements in mortality
 investment return
 salary growth
 pension increases
 decrement rates, eg rates of withdrawal or ill-health
 option take-up rates, eg early retirement

(ii) Why different schemes may use different assumptions

Purpose of the valuation

Different schemes may have different funding targets, eg some schemes

may be aiming to set aside sufficient funds to buy annuities with an insurer.

© IFE: 2019 Examinations Page 137

 Exclusive use

Needs of the client

The views of trustees may differ, some may wish to be more prudent.

Actuarial advisors may have different views, eg different views on investment

returns or longevity.

External influences

Guidance from the regulator on the assumptions to be used may have been
interpreted differently.

There may be different legal views on the benefits to be valued.

Scheme specifics

The value of an assumption may be scheme-specific, eg salary growth.

Profile of membership

Membership profiles may be different, eg one scheme may be more mature.

The demographic characteristics of members (eg longevity) may be different

due to differences in location, occupation and socio-economic group.

There may be a large number of members with small liabilities or a small

number of members with large liabilities.

Different assumptions and methods may be needed for open and closed
schemes.

For closed schemes, the membership will age over time, but an open
scheme may have a fairly stable population (if new entrants replace leavers).

Strength of sponsor

Different schemes will have sponsors with different levels of financial

strength.

Schemes with strong sponsors may be able to adopt a less prudent

approach.

Investment strategy

Different schemes may have different investment strategies.

Page 138 © IFE: 2019 Examinations

 Exclusive use

A well-matched investment strategy may reduce the importance of specific

assumptions, eg investment returns and inflation.

A prudent investment strategy involving mainly bonds (compared with a

scheme invested primarily in equities) could have different assumptions, as
lower expected returns on assets may be assumed.

Different bonds, eg with different credit ratings or different terms, may have
been used to assess bond yields.

Asset valuation

The methods chosen to value assets may differ, eg schemes could use
market value or discounted cashflow approaches.

Size of scheme

Schemes vary in size, leading to different assumptions being used.

For example, scheme size may affect the choice of assets and hence
investment returns.

Benefits valued

Different benefits may be valued, eg dependants’ benefits or discretionary

benefits.

Different measures of inflation may be used in the scheme rules for benefit
increases.

Some schemes have caps and floors on benefit increases.

Different schemes may offer different options and guarantees, eg some

schemes may offer a tax-free lump sum which may have a different value
than the pension given up, and so the take up rate of the option is important.

The assumption for the take-up rate may differ depending on the purpose of
the valuation and the level of prudence required.

Different approaches may have been taken to insured benefits.

Data used

Some schemes may have more extensive data, eg about members’ health.

© IFE: 2019 Examinations Page 139

 Exclusive use

Assumptions may be affected by the quality of data.

Assumptions may be set using scheme-specific data or standard data.

Modelling approach

Different models may have been used to assess the risks, eg schemes could
use their own models or models published by the actuarial profession.

Different schemes may take a more, or less, detailed approach to modelling.

Funding level

Schemes that are well funded may take a more relaxed approach to funding,
meaning less prudent assumptions are used.

Funding approach

Different funding approaches may be used, eg most schemes will fund in

advance, but it may be that regulation permits the use of other methods such
as pay-as-you-go or just-in-time funding.

Scheme expenses

Scheme expenses may differ, meaning different assumptions are made,

eg depending on whether they are paid by the scheme or by the sponsor.

43 Subject CA1 Paper 1 September 2017 Question 3

(i) Areas of actuarial work where data is required

 statutory returns
 accounts
 investment management
 risk management
 management information / financial control
 pricing
 experience statistics / analyses
 marketing
 provisioning

Page 140 © IFE: 2019 Examinations

 Exclusive use

(ii) Mitigating and making allowances for poor quality data

General

Problems with data may arise from:

 poor quality data
 an insufficient quantity of data
 the data not being sufficiently detailed
 data that is not relevant.

Data issues are best resolved by getting to the root of the problem.

Data issues should be fully disclosed to the client.

Warning should be given about the extent to which the results can be relied
upon.

Mitigation of data problems

Detail

 Data can be improved in terms of detail by ensuring the inputs to the

system (eg proposal form) capture all the detail required.
 It may be necessary to use summarised data, but it can be more difficult
to validate.

Quality

 If data has been supplied by third parties then it will need to be verified.
 The proposal form should ask clear unambiguous questions to reduce
the risk of data error.
 Automated data validation should be carried out, eg checking for blank
fields or impossible dates.
 The data should be reconciled with previous years.
 Checks should be carried out that the data is consistent and has no
obvious errors.
 Spot checks should be carried out on records.

© IFE: 2019 Examinations Page 141

 Exclusive use

Quantity

 Data can be supplemented with external data, although care needs to

be taken to check the data is relevant.

Relevance

 Data needs to be collated into homogeneous groups.

 This needs to be balanced against the need for credibility.

Making allowances for poor data

Quality

 ‘Making allowances’ means that the data has been used to produce a
result where that data may be sub-optimal, and so the results may not
be reliable.
 Risk margins or contingency loadings may need to be included; the size
of any margins or loadings will be based on actuarial judgement.
 Independent checks of the reasonableness of the results should be
carried out, which may highlight data issues.
 Comparisons may be carried out against the results of competitors to
see if the results are consistent.
 Sensitivity testing of data quality can be carried out.

44 Subject CA1 Paper 1 September 2017 Question 6 (part)

Why inappropriate to change assumptions

Relevance of the data

The data may be distorted by random fluctuations or abnormal fluctuations.

The source of data lacks credibility, since the data is only ‘anecdotal’ …

… need to consider the underlying source.

Future trends need to be allowed for …

… and may not have been adequately reflected in the data.

Page 142 © IFE: 2019 Examinations

 Exclusive use

Certain groups may be under- or over- represented in the data …

… eg if the change only affects lower income females who are

over-represented in the data.

Credibility of the data

Three years is too short a period of time to consider …

… particularly for assessing mortality experience.

There is likely to be insufficient data to support a change in rates.

Implications of using the data for pricing etc

Changing the assumptions is a significant exercise, eg if other insurers do

not change their rates the insurer’s premium rates will be unattractive …

… it is therefore important that such changes are based on reliable data.

A more detailed investigation should be conducted …

… before data is used for pricing or reserving.

Assumptions are adjusted to reflect expected future experience …

… over the term of the contract, which can be very long.

A more detailed investigation would be needed to determine age-specific

rates.

45 Subject CA1 Paper 1 September 2017 Question 7 (part)

(i) Basic features of general actuarial models used to project future

cashflows

An actuarial model needs to allow for all the cashflows that may arise.

These will depend on the nature of the financial products, schemes,

contracts or transactions being modelled …

… and any discretionary benefits they carry.

The cashflows need to allow for any interactions …

… particularly where assets and liabilities are being modelled together.

© IFE: 2019 Examinations Page 143

 Exclusive use

In some cases there is a need to use stochastic models and simulation.

The time period for calculating the cashflows in the projection needs to be
chosen …

… bearing in mind that the more frequently the cashflows are calculated the
more reliable the output from the model …

…. although there is a danger of spurious accuracy …

… the less frequently the cashflows are calculated the faster the model can
be run and results obtained.

Allowance should be made, where appropriate, for the cashflows arising

from any need to hold reserves …

… and to maintain spare capital.

Where the business being modelled includes options, the potential

cashflows from such options …

… and the take-up rate need to be allowed for.

(ii) How sensitivity testing would be used when modelling the

cashflows

The model will initially be run using a best-estimate basis, but there will be
significant uncertainty around some of the assumptions.

So the model will need to be re-run …

… using a range of feasible assumptions.

Sensitivity testing involves changing assumptions one at a time.

When an assumption is changed, other assumptions which are correlated

with it should also be changed.

Profitability will be particularly sensitive to the revenue from ticket sales so

this will be a key assumption to vary.

The exchange rates are also an important assumption to test …

… given many of the cashflows will be in overseas currencies.

Page 144 © IFE: 2019 Examinations

 Exclusive use

The future revenue stream is likely to be correlated with the success of the
tour, eg ticket sales from this tour …

… as a successful tour will lead to more valuable future sponsorship and …

… increased future ticket sales, CD and other merchandise sales.

Precise correlations may however be difficult to gauge and these

correlations should also be tested.

The most important assumptions will relate to the probabilities underlying

each expected cashflow …

… ie how likely different levels of revenue and expenditure are …

… so re-runs can be carried out varying these probabilities.

The expenditure on the tour is likely to be more predictable, given many

costs will be pre-paid.

Expenditure will be less predictable if insurance has not been used …

… ie there may be unexpected costs arising from risk events which need to
be met by the orchestra.

The assessment of exchange rates can allow for the correlation between
different exchange rates.

As this is a three-week tour the discount rate is not an important assumption,

although if future revenue and expenditure after the tour is included it
becomes more important.

There is also merit in looking at different scenarios by changing many

assumptions in combination.

© IFE: 2019 Examinations Page 145

 Exclusive use

Page 146 © IFE: 2019 Examinations

 Exclusive use

BRAINSTORM (2)

Hopefully, by now, you will have attempted the majority of the Core Reading
and Exam Questions. Have another go at brainstorming the topic of
modelling. You should be able to come up with more ideas, more quickly
than before.

Requirements

Modelling

Uses
Steps

Cashflows

© IFE: 2019 Examinations Page 147

 Exclusive use

BRAINSTORM – SOLUTION

This is an example of the brainstorm that you should have produced. It is

not supposed to be exhaustive and there may be other areas that you have
covered.

Reflect expected
Special Parameter values
experience
features appropriate
Independently
verifiable
Dynamic
Parameters for each
Outputs significant feature
Easily Margins
communicated
Reflects risk profile of
business Discount rate
Not too
Easy to
complex
interpret
Requirements Well documented Scenario
test/stochastic?
Not too time
consuming
Rigorous

Valid Marketability?
Not too expensive
Profit criteria?
Deterministic Capital?
Cashflow
ALM
Speed vs reliability
Objective Solvency?
Variability
Types

Correlations
Modelling Time horizon/period Data
Stochastic
New/existing

Model points
Cost guarantees Solvency
Uses
Options/
Steps Parameters
Pricing guarantees
Expenses
Variables
Financing
Cashflows
strategy ALM

Goodness of fit
Run
Premiums/ Senstivity
contributions test
Profitability
Project/discount
cashflows
Risk management / Benefits Investment
Output
capital income/gains

Page 148 © IFE: 2019 Examinations

 Exclusive use

FINAL COMMENTS

As well as working on your question answering skills, here are the key things
from this booklet that you should be learning for the exam (you might want to
tick them off when you have done them):
 the operational issues that need to be considered when designing a
model (see Acronym 1 below)
 the cashflows to include in a model
 the pros and cons of deterministic / stochastic models
 the factors to consider when deciding what model to use (see
Acronym 2 below)
 the steps in running a model and the application of these to various
situations – pricing, setting benefit scheme financing strategies,
assessing capital requirements, costing guarantees
 ways of assessing the variability of experience (sensitivity testing,
scenario testing and stochastic modelling)
 the different types of personal data and the nature of ‘big data’,
including the related ethical / regulatory issues and data governance
considerations
 the sources of data (see Acronym 3 below)
 problems relating to data quality (see Acronym 4 below) and how to
overcome them
 data checks
 reasons for heterogeneity in industry-wide data
 a description of risk classification
 the main considerations in choosing assumptions (see Acronym 5
below)
 considerations when using past data to set future assumptions (see
Acronym 6 below)
 examples of historical and current data that can be used to help setting
assumptions
 why this data may need adjusting to be suitable for setting future
assumptions

© IFE: 2019 Examinations Page 149

 Exclusive use

 how assumptions need to reflect the objectives of the clients in different

situations
 ways of allowing for risk in assumptions
 features that increase the riskiness of a contract (and so the risk
discount rate)
 examples of profit criteria.

Acronym 1 – Model design: operational issues

Simple but retains key features

Clear results
Adequately documented
Range of implementation methods
Communicable workings and outputs
Easy to understand
Refineable & developable

Frequency of cashflows – balance accuracy vs practicality

Independent verification of outputs
Length of run not too long
Expense not too high
Sensible joint behaviour of variables

Acronym 2 – Considerations in assessing different models

Fit for purpose

Expertise available in house
Need flexibility
Cost of each option
Expected number of times used
Desired accuracy

Acronym 3 – Sources of data

Tables, eg actuarial mortality tables

Reinsurers
Abroad (data from overseas contracts)
Industry data
National statistics
Experience investigations on the existing contract
Regulatory reports and company accounts
Similar contracts

Page 150 © IFE: 2019 Examinations

 Exclusive use

Acronym 4 – Potential issues when using data

Quantity (credibility)
Up-to-date?
Errors
Relevance (heterogeneity)
Incomplete?
Exceptionals
Detail insufficient

Acronym 5 – Factors to consider when setting assumptions

Legislation / regulation
Use of the data
Needs of the client
Consistency between assumptions
How financially significant is/are the assumption(s)

Acronym 6 – Considerations when using past data to set future assumptions

Balance of homogenous groups underlying the data may have changed

Economic situation may have changed
Social conditions may have changed
Trends over time, eg medical, demographic

Abnormal fluctuations
Random fluctuations
Changes in regulation
Heterogeneity within the group to which the assumptions will apply
Errors in data
Recording differences (eg in categorisation of smoker)

We hope that you have found this booklet helpful with your revision. If you
are able to give us any feedback to help us improve future editions of this
booklet then please email your comments to [email protected].

© IFE: 2019 Examinations Page 151

 Exclusive use

NOTES

Page 152 © IFE: 2019 Examinations

 Exclusive use

NOTES

© IFE: 2019 Examinations Page 153

 Exclusive use

NOTES

Page 154 © IFE: 2019 Examinations

 Exclusive use

NOTES

© IFE: 2019 Examinations Page 155

 Exclusive use

NOTES

Page 156 © IFE: 2019 Examinations

 Exclusive use

NOTES

© IFE: 2019 Examinations Page 157

 Exclusive use

EXAM PREPARATION CHECKLIST

Core Reading Questions

We recommend you work through the Core Reading Questions at least three
times before the exam. Frequent review leads to quick and reliable recall.

Date of first attempt: ___________________

Date of second attempt: ___________________

Date of third attempt: ___________________

Past Exam Questions

We recommend that you work through as many questions as possible under

exam conditions. By doing so, you’ll get used to the style of answers
required and to working under time pressure. Keep a note of which
questions you have attempted and when.

Qn Date Qn Date Qn Date

attempted attempted attempted
1 16 31
2 17 32
3 18 33
4 19 34
5 20 35
6 21 36
7 22 37
8 23 38
9 24 39
10 25 40
11 26 41
12 27 42
13 28 43
14 29 44
15 30 45

Page 158 © IFE: 2019 Examinations