Modelling Mortality with Actuarial Applications

Modern mortality modelling for actuaries and actuarial students, with example R code, to unlock the potential of individual data.

Angus S. Macdonald (Author), Stephen J. Richards (Author), Iain D. Currie (Author)

9781107045415, Cambridge University Press

Hardback, published 3 May 2018

384 pages, 95 b/w illus. 42 tables
23.5 x 15.8 x 2 cm, 0.74 kg

'A wonderfully comprehensive piece of work from leading experts in the field. Essential reading for anyone with an interest in longevity modelling.' Richard Willets, ReAssure Ltd.

Actuaries have access to a wealth of individual data in pension and insurance portfolios, but rarely use its full potential. This book will pave the way, from methods using aggregate counts to modern developments in survival analysis. Based on the fundamental concept of the hazard rate, Part I shows how and why to build statistical models, based on data at the level of the individual persons in a pension scheme or life insurance portfolio. Extensive use is made of the R statistics package. Smooth models, including regression and spline models in one and two dimensions, are covered in depth in Part II. Finally, Part III uses multiple-state models to extend survival models beyond the simple life/death setting, and includes a brief introduction to the modern counting process approach. Practising actuaries will find this book indispensable, and students will find it helpful when preparing for their professional examinations.

Preface
Part I. Analysing Portfolio Mortality: 1. Introduction
2. Data preparation
3. The basic mathematical model
4. Statistical inference with mortality data
5. Fitting a parametric survival model
6. Model comparison and tests of fit
7. Modelling features of the portfolio
8. Non-parametric methods
9. Regulation
Part II. Regression and Projection Models: 10. Methods of graduation I – regression models
11. Methods of graduation II – smooth models
12. Methods of graduation III – 2-dimensional models
13. Methods of graduation IV – forecasting
Part III. Multiple-State Models: 14. Markov multiple-state models
15. Inference in the Markov model
16. Competing risks models
17. Counting-process models
Appendix A. R commands
Appendix B. Basic likelihood theory
Appendix C. Conversion to published tables
Appendix D. Numerical integration
Appendix E. Mean and variance-covariance of a vector
Appendix F. Differentiation with respect to a vector
Appendix G. Kronecker product of two matrices
Appendix H. R functions and programs
References
Author index
Index.

Subject Areas: Optimization [PBU], Insurance & actuarial studies [KFFN], Finance [KFF], Economic statistics [KCHS], Risk assessment [GPQD]