×

Negative binomial version of the Lee-Carter model for mortality forecasting. (English) Zbl 1150.91426

The authors assume that the force of mortality at age \(x\) in calendar year \(t\) is of the form \(\exp(\alpha_{x}+\beta_{x}\kappa_{t})\). Interpretation of the parameters is the following: \(\exp\alpha_{x}\) is the general shape of the mortality schedule and the actual forces of mortality change according to an overall mortality index \(\kappa_{t}\) modulated by an age response \(\beta_{x}\). This paper develops a negative binomial regression model for estimating the \(\alpha_{x},\beta_{x}, \kappa_{t}\). The parameters are estimated by maximum likelihood. A numerical illustration with Swedish, French and Italian data, general population is presented. The overdispersion present in population data is detected using the test statistics proposed by A. C. Cameron and P. K. Trivedi [J. Applied Econometrics, 46, 347-364 (1986)]. The goodness of fit obtained in the negative binomial regression model for death counts is examined and a comparison with the Poisson model is performed using information criteria.

MSC:

91B30 Risk theory, insurance (MSC2010)
PDF BibTeX XML Cite
Full Text: DOI

References:

[1] Lee, Journal of the American Statistical Associations 87 pp 659– (1992)
[2] Lee, North American Actuarial Journal 4 pp 80– (2000) · Zbl 1083.62535
[3] Booth, International Journal of Forecasting 22 pp 547– (2006)
[4] Brouhns, Insurance: Mathematics and Economics 31 pp 373– (2002)
[5] Brouhns, Bulletin of the Swiss Association of Actuaries 2 pp 105– (2002)
[6] Czado, Insurance: Mathematics and Economics 36 pp 260– (2005)
[7] Cossette, North American Actuarial Journal (2007)
[8] Brown, Journal of Risk and Insurance 70 pp 17– (2003)
[9] Cameron, Journal of Applied Econometrics 46 pp 347– (1986) · Zbl 04500089
[10] Goodman, Journal of the American Statistical Association 74 pp 537– (1979)
[11] Kuha, Sociological Methods and Research 33 pp 188– (2004)
[12] . Model Selection and Multi-model Inference: A Practical Information–Theoretic Approach. Springer: New York, 2002.
[13] Bayesian model selection in social research. In Sociological Methodology, (ed.). The American Sociological Association: Washington, 1995; 111–163.
[14] Renshaw, Journal of the Institute of Actuaries 119 pp 69– (1992)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.