Bayesian Poisson log-bilinear mortality projections

Claudia Czado; Antoine Delwarde; Michel Denuit

doi:10.1016/j.insmatheco.2005.01.001

Bayesian Poisson log-bilinear mortality projections

Claudia Czado, Antoine Delwarde, Michel Denuit

Associate Professorship of Applied Mathematical Statistics

Centre Hospitalier Universitaire (CHU) Mont-Godinne

Research output: Contribution to journal › Article › peer-review

106 Scopus citations

Abstract

Mortality projections are major concerns for public policy, social security and private insurance. This paper implements a Bayesian log-bilinear Poisson regression model to forecast mortality. Computations are carried out using Markov Chain Monte Carlo methods in which the degree of smoothing is learnt from the data. Comparisons are made with the approach proposed by Brouhns et al. [Insur.: Math. Econ. 31 (2002) 373-393; Bull. Swiss Assoc. Actuaries (2002) 105-130], as well as with the original model of Lee and Carter [J. Am. Stat. Assoc. 87 (1992) 659-671].

Original language	English
Pages (from-to)	260-284
Number of pages	25
Journal	Insurance: Mathematics and Economics
Volume	36
Issue number	3
DOIs	https://doi.org/10.1016/j.insmatheco.2005.01.001
State	Published - 24 Jun 2005

Keywords

Expected remaining lifetimes
MCMC
Poisson regression
Projected lifetables

Access to Document

10.1016/j.insmatheco.2005.01.001

Cite this

@article{86c48cb0eb6d4d6fb7bc408aeaad9632,

title = "Bayesian Poisson log-bilinear mortality projections",

abstract = "Mortality projections are major concerns for public policy, social security and private insurance. This paper implements a Bayesian log-bilinear Poisson regression model to forecast mortality. Computations are carried out using Markov Chain Monte Carlo methods in which the degree of smoothing is learnt from the data. Comparisons are made with the approach proposed by Brouhns et al. [Insur.: Math. Econ. 31 (2002) 373-393; Bull. Swiss Assoc. Actuaries (2002) 105-130], as well as with the original model of Lee and Carter [J. Am. Stat. Assoc. 87 (1992) 659-671].",

keywords = "Expected remaining lifetimes, MCMC, Poisson regression, Projected lifetables",

author = "Claudia Czado and Antoine Delwarde and Michel Denuit",

note = "Funding Information: Claudia Czado greatly acknowledges the support through the SFB 386 “Statistical Analysis of discrete Structures” sponsored by the Deutsche Forschungsgemeinschaft. Funding Information: Antoine Delwarde was supported by a research grant from the Fonds de la Recherche pour l{\textquoteright}Industrie et l{\textquoteright}Agriculture- FRIA. This financial support is gratefully acknowledged. ",

year = "2005",

month = jun,

day = "24",

doi = "10.1016/j.insmatheco.2005.01.001",

language = "English",

volume = "36",

pages = "260--284",

journal = "Insurance: Mathematics and Economics",

issn = "0167-6687",

publisher = "Elsevier B.V.",

number = "3",

}

TY - JOUR

T1 - Bayesian Poisson log-bilinear mortality projections

AU - Czado, Claudia

AU - Delwarde, Antoine

AU - Denuit, Michel

N1 - Funding Information: Claudia Czado greatly acknowledges the support through the SFB 386 “Statistical Analysis of discrete Structures” sponsored by the Deutsche Forschungsgemeinschaft. Funding Information: Antoine Delwarde was supported by a research grant from the Fonds de la Recherche pour l’Industrie et l’Agriculture- FRIA. This financial support is gratefully acknowledged.

PY - 2005/6/24

Y1 - 2005/6/24

N2 - Mortality projections are major concerns for public policy, social security and private insurance. This paper implements a Bayesian log-bilinear Poisson regression model to forecast mortality. Computations are carried out using Markov Chain Monte Carlo methods in which the degree of smoothing is learnt from the data. Comparisons are made with the approach proposed by Brouhns et al. [Insur.: Math. Econ. 31 (2002) 373-393; Bull. Swiss Assoc. Actuaries (2002) 105-130], as well as with the original model of Lee and Carter [J. Am. Stat. Assoc. 87 (1992) 659-671].

AB - Mortality projections are major concerns for public policy, social security and private insurance. This paper implements a Bayesian log-bilinear Poisson regression model to forecast mortality. Computations are carried out using Markov Chain Monte Carlo methods in which the degree of smoothing is learnt from the data. Comparisons are made with the approach proposed by Brouhns et al. [Insur.: Math. Econ. 31 (2002) 373-393; Bull. Swiss Assoc. Actuaries (2002) 105-130], as well as with the original model of Lee and Carter [J. Am. Stat. Assoc. 87 (1992) 659-671].

KW - Expected remaining lifetimes

KW - MCMC

KW - Poisson regression

KW - Projected lifetables

UR - http://www.scopus.com/inward/record.url?scp=21244478558&partnerID=8YFLogxK

U2 - 10.1016/j.insmatheco.2005.01.001

DO - 10.1016/j.insmatheco.2005.01.001

M3 - Article

AN - SCOPUS:21244478558

SN - 0167-6687

VL - 36

SP - 260

EP - 284

JO - Insurance: Mathematics and Economics

JF - Insurance: Mathematics and Economics

IS - 3

ER -

Bayesian Poisson log-bilinear mortality projections

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this