Multiple Decrement Models in Insurance

Shailaja Deshmukh

Multiple Decrement
Models in Insurance

An Introduction Using R
Shailaja Deshmukh
Department of Statistics
University of Pune
Pune, India

ISBN 978-81-322-0658-3 ISBN 978-81-322-0659-0 (eBook)

DOI 10.1007/978-81-322-0659-0
Springer New Delhi Heidelberg New York Dordrecht London

Library of Congress Control Number: 2012942476

© Springer India 2012

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered
and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of
this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from Springer.
Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations
are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of pub-
lication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any
errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect
to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Preface

Actuarial profession is one of the prestigious professions around the world. An actu-
ary has to evaluate the entire operation of the insurance business deploying a variety
of tools from actuarial science, which are heavily based on the statistical techniques
and principles from finance and economics. With the liberalization of the insurance
industry in India, the demand for actuaries and actuarial courses is increasing. The
aim of this book is to cater to the needs of students intending to pursue actuarial pro-
fession and to familiarize them with application of various statistical methods used
in the insurance industry. The book elaborates on actuarial concepts and statistical
techniques in multiple decrement models with their application in pension funding,
and multi-state transition models with application in disability income insurance.
This book is written in the same style as my book “Actuarial Statistics: An Intro-
duction Using R” published by Universities Press, India, in 2009, which discusses
statistical tools for the computations of premiums and reserves for life insurance
products and annuities in single decrement models, using R software.
In some policies, benefit to a single life or a group, is subject to a type of contin-
gency. For example, the death of an individual may be due to an accident or due to
any other cause. In most of the insurance policies, the coverage is firstly given for
the base cause, and then there are policy riders for additional benefits. If death is due
to an accident, then the benefit structure is different; usually benefit is more than the
base coverage. In such cases, the benefit structure and consequently the premium
structure depend on time to death and the cause of the death. Survivorship models
incorporating two random mechanisms, time to termination and various modes of
termination, are known as multiple decrement models. The first chapter introduces
the multiple decrement model and the construction of the multiple decrement table
using the associated single decrement model and central rate bridge.
Chapter 2 discusses calculation of premiums and reserves in life insurance prod-
ucts when the benefit depends on the cause of decrement along with the time to
decrement.
A major application of multiple decrement models is in pension plans and em-
ployee benefit plans. In these schemes, the benefit paid on termination of employ-
ment depends upon the several causes of termination. The cause of termination may

v
vi Preface

be withdrawal, disability, death, or retirement. The benefits on retirement often dif-

fer from those payable on death or disability. As a consequence, the actuarial present
value of the benefits depends on the cause of death along with the future life time
of an individual. To determine the rate of contribution in pension funds and to value
the pension fund at specified times, it is necessary to find the actuarial present value
of the benefits. Therefore, survivorship models for employee benefit systems and
pension funds include random variables for both time to termination and cause of
termination. Chapter 3 is devoted to the application of multiple decrement models
for evaluating the cost of a given pension plan at a specific time. Once the estimate
of the ultimate cost of the plan is determined, the next step is to determine the con-
tributions required to pay for the estimated cost in an orderly manner, so that the
estimated cost of the plan is spread over future years. These actuarial techniques
are referred to as actuarial cost methods or actuarial funding methods. A funding
method specifies the pattern, that is, the frequency, and the amount of aggregate
contributions required to balance the benefit payments. Chapter 4 reviews some of
these methods.
As an extension of multiple decrement models, the multi-state model of transi-
tions is discussed in Chap. 5, when the transitions among the states are governed
by Markov models. Multiple state models have proved to be appropriate models
for an insurance policy in which the payment of benefits or premiums is dependent
on being in a given state or moving between a given pair of states at a given time.
Such a model is useful to decide premium in continuing care retirement communi-
ties model in health insurance and disability income insurance model in employee
benefit schemes.
In all these chapters, it is assumed that the rate of interest in the calculations
of actuarial present values is deterministic and usually constant over the period of
policy. However, the assumption of deterministic interest will be rarely realized in
practice, particularly for long-term policies. Chapter 6 introduces in brief stochastic
models for interest rates and calculation of premiums for some products in this
setup.
The highlight of the book is its usage of R software for statistical computations.
R software is freely available from public domain. In all the Universities in India and
abroad the use of R software is increasing tremendously. Most of the recent books
incorporate R software for statistical analysis. To be consistent with the recent trend
and demand, R software is used in this book to compute various monetary functions
involved in insurance business. R commands are given for all the computations, and
meanings of these are explained, so that a reader unfamiliar with R can also use
it. All the tables inserted in the book and solutions to all illustrative examples are
worked out using R. The command-driven R software brings out very clearly the
successive stages in statistical computations.
The book builds on from the very basic concepts, defining and explaining the
terms and to move on to their applications and actual computations with R. It is easy
to follow and moves on step-by-step from basics to detailed calculations. The book
contains many solved examples illustrating the theory. For better assimilation of the
material, exercises are given at the end of each chapter. Statistical prerequisites to
Preface vii

the book are concepts and computation of premiums and reserves for some standard
insurance products based on single decrement models.
I hope that this book will be instructive and will induce interest among the stu-
dents about the actuarial profession. I am sure the book will be helpful for those
who wish to prepare for examinations conducted by actuarial societies worldwide.
Feedback, in the form of suggestions and comments from colleagues, students,
and all readers, is most welcome.
I thank all my friends, colleagues, and family members for encouragement and
support received throughout this venture. I am indeed thankful to the students who
opted for this course in the last couple of years. They provided me the incentive to
study rigorously and to collect and set a variety of problems, all of which are helpful
in writing this book.
Pune, India S. Deshmukh
Biography

Shailaja Deshmukh is a Professor of Statistics at the University of Pune, India. Her

areas of interest are inference in stochastic processes, applied probability and anal-
ysis of microarray data. She has authored three books—Microarray Data: Statistical
Analysis Using R (jointly with Dr. Sudha Purohit), Statistics Using R (jointly with
Dr. Sudha Purohit and Prof. Sharad Gore) and Actuarial Statistics: An Introduc-
tion Using R. She has a number of research publications in various peer-reviewed
journals.

ix
Contents

1 Multiple Decrement Models . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Time to Decrement and Cause of Decrement Random Variables . . 3
1.3 Multiple Decrement Table . . . . . . . . . . . . . . . . . . . . . . 22
1.4 Associated Single Decrement Model . . . . . . . . . . . . . . . . 29
1.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2 Premiums and Reserves in Multiple Decrement Model . . . . . . . . 53
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2 Actuarial Present Value of Benefit . . . . . . . . . . . . . . . . . . 54
2.3 Computation of Premiums . . . . . . . . . . . . . . . . . . . . . . 58
2.4 Computation of Reserves . . . . . . . . . . . . . . . . . . . . . . 75
2.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3 Defined Benefit Pension Plan . . . . . . . . . . . . . . . . . . . . . . 81
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.2 Actuarial Present Value of Pension Benefit . . . . . . . . . . . . . 84
3.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
4 Pension Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.2 Accrued Benefit Cost Method for an Individual . . . . . . . . . . . 132
4.3 Accrued Benefit Cost Method for a Group . . . . . . . . . . . . . 143
4.4 Aggregate Actuarial Cost Method . . . . . . . . . . . . . . . . . . 163
4.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5 Multi-state Transition Models for Cash Flows . . . . . . . . . . . . . 173
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.2 Markov Chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
5.3 Actuarial Present Values of Cash Flows . . . . . . . . . . . . . . . 179
5.4 Markov Process Model . . . . . . . . . . . . . . . . . . . . . . . 189
5.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

xi
xii Contents

6 Stochastic Interest Rate . . . . . . . . . . . . . . . . . . . . . . . . . 205

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
6.2 Random Interest Scenario . . . . . . . . . . . . . . . . . . . . . . 205
6.3 Parametric Models . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6.4 Time Series Models . . . . . . . . . . . . . . . . . . . . . . . . . 212
6.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
List of Figures

(τ )
Fig. 1.1 Graphs of t px for x = 30, 40, 50, and 60 . . . . . . . . . . . . . . 7
Fig. 1.2 Graphs of g(t), f (t, 1), and f (t, 2) . . . . . . . . . . . . . . . . . 8
Fig. 1.3 Graph of survival function of T (x) . . . . . . . . . . . . . . . . . 15
Fig. 1.4 Graph of probability density function of T (x) . . . . . . . . . . . . 16
Fig. 1.5 Forces of decrement . . . . . . . . . . . . . . . . . . . . . . . . . 38
Fig. 4.1 Stationary and stable population with R = 0.02 . . . . . . . . . . . 150
Fig. 4.2 Stationary and stable population with R = −0.01 . . . . . . . . . . 150

xiii
List of Tables

Table 1.1 The output, formatted with the help of Excel . . . . . . . . . . . 12

Table 1.2 Survival probability . . . . . . . . . . . . . . . . . . . . . . . . 15
Table 1.3 Table of decrement probabilities . . . . . . . . . . . . . . . . . 16
Table 1.4 Probability function of J (x) . . . . . . . . . . . . . . . . . . . 17
Table 1.5 Conditional distribution of J (x) given T (x) = 10 . . . . . . . . 17
Table 1.6 Joint distribution of K(30) and J (30); Marginal distribution
of K(30) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Table 1.7 Decrement probabilities . . . . . . . . . . . . . . . . . . . . . . 25
Table 1.8 Expected number of decrements . . . . . . . . . . . . . . . . . 27
Table 1.9 Expected number of decrements . . . . . . . . . . . . . . . . . 27
Table 1.10 Decrement and survival probabilities . . . . . . . . . . . . . . . 28
Table 1.11 Double decrement table . . . . . . . . . . . . . . . . . . . . . . 29
Table 1.12 The extract from a triple-decrement table . . . . . . . . . . . . . 32
Table 1.13 Associated single decrement probabilities . . . . . . . . . . . . 36
Table 1.14 Multiple decrement table . . . . . . . . . . . . . . . . . . . . . 36
Table 1.15 Multiple decrement table from associated single decrement model 38
Table 1.16 Values of associated single decrement rates . . . . . . . . . . . . 39
Table 1.17 Constant for each decrement over the year of age 35, 36, and 37,
(1) (2)
q35 = 0.01, q35 = 0.06, and further information . . . . . . . . 40
Table 1.18 The data for a double-decrement model . . . . . . . . . . . . . . 41
Table 1.19 The absolute rates . . . . . . . . . . . . . . . . . . . . . . . . . 46
Table 1.20 Multiple decrement probabilities . . . . . . . . . . . . . . . . . 48
Table 1.21 Decrement probabilities . . . . . . . . . . . . . . . . . . . . . . 49
Table 1.22 Decrement probabilities . . . . . . . . . . . . . . . . . . . . . . 50
Table 1.23 The number of survivors and number of deaths due to two causes 50
Table 1.24 The basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Table 1.25 The force of decrement is constant for each decrement over
(1) (2)
the year of age 45, 46, and 47, q45 = 0.02, q45 = 0.08,
and further information . . . . . . . . . . . . . . . . . . . . . . 51
Table 1.26 Absolute rates . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Table 2.1 Premium for whole life insurance . . . . . . . . . . . . . . . . . 64

xv
xvi List of Tables

Table 2.2 n-year premium payment . . . . . . . . . . . . . . . . . . . . . 65

Table 2.3 Premium for n-year term insurance . . . . . . . . . . . . . . . . 65
Table 2.4 Premiums for varying benefits . . . . . . . . . . . . . . . . . . 68
Table 2.5 n-year premium . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Table 2.6 Premium for n-year term insurance . . . . . . . . . . . . . . . . 70
Table 2.7 n-year discrete premiums for whole life insurance . . . . . . . . 73
Table 2.8 Discrete premium for n-year term insurance . . . . . . . . . . . 74
Table 2.9 Discrete reserves for a policy with rider . . . . . . . . . . . . . 79
Table 3.1 Decrement and survival probabilities . . . . . . . . . . . . . . . 97
Table 3.2 Multiple decrement table . . . . . . . . . . . . . . . . . . . . . 98
Table 3.3 Service table . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Table 3.4 Multiple decrement table . . . . . . . . . . . . . . . . . . . . . 102
Table 3.5 Joint and marginal distributions of K and J . . . . . . . . . . . 105
Table 3.6 Estimated salary . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Table 3.7 Projected annual benefit at retirement . . . . . . . . . . . . . . . 111
Table 3.8 Discrete monthly life annuity . . . . . . . . . . . . . . . . . . . 119
Table 3.9 Predicted annual salary and contribution . . . . . . . . . . . . . 127
Table 4.1 Accrued actuarial liability . . . . . . . . . . . . . . . . . . . . . 138
Table 4.2 Age wise normal cost rate and accrued actuarial liability . . . . . 141
Table 4.3 Population size with R = 0 and R = 0.02 . . . . . . . . . . . . . 148
Table 4.4 Population size with R = 0 and R = −0.01 . . . . . . . . . . . 149
Table 4.5 Total payroll payment rate . . . . . . . . . . . . . . . . . . . . 151
Table 4.6 Normal cost rate in terminal funding . . . . . . . . . . . . . . . 152
Table 4.7 Terminal fund with initial benefit b1 (65) . . . . . . . . . . . . . 153
Table 4.8 Terminal fund with initial benefit b2 (65) . . . . . . . . . . . . . 153
Table 4.9 Normal cost rate with m1 (x) . . . . . . . . . . . . . . . . . . . 161
Table 4.10 Normal cost rate with m2 (x) . . . . . . . . . . . . . . . . . . . 161
Table 4.11 Accrued actuarial liability with m1 (x) . . . . . . . . . . . . . . 161
Table 4.12 Accrued actuarial liability with m2 (x) . . . . . . . . . . . . . . 162
Table 4.13 Normal cost with initial benefit 300000 with m1 (x) . . . . . . . 162
Table 4.14 Normal cost with initial benefit 300000 with m2 (x) . . . . . . . 162
Table 4.15 Amortization values with m1 (x) . . . . . . . . . . . . . . . . . 166
Table 4.16 Amortization values with m2 (x) . . . . . . . . . . . . . . . . . 166
Table 4.17 Contribution rate for m1 (x) . . . . . . . . . . . . . . . . . . . . 170
Table 4.18 Contribution rate for m2 (x) . . . . . . . . . . . . . . . . . . . . 170
Table 4.19 Compensation for unfunded liability for m1 (x) . . . . . . . . . . 171
Table 4.20 Compensation for unfunded liability for m2 (x) . . . . . . . . . . 171
Table 6.1 Interest rate scenario . . . . . . . . . . . . . . . . . . . . . . . 207
Table 6.2 Actuarial present values . . . . . . . . . . . . . . . . . . . . . . 208
Table 6.3 Interest rate scenario . . . . . . . . . . . . . . . . . . . . . . . 215