A hierarchical model for the joint mortality analysis of pension scheme data with missing covariates

Research output: Contribution to journalArticle

Abstract

A hierarchical model is developed for the joint mortality analysis of pension scheme datasets. The proposed model allows for a rigorous statistical treatment of missing data. While our approach works for any missing data pattern, we are particularly interested in a scenario where some covariates are observed for members of one pension scheme but not the other. Therefore, our approach allows for the joint modelling of data sets which contain
different information about individual lives. The proposed model generalizes the specification of parametric models when accounting for covariates.

We consider parameter uncertainty using Bayesian techniques. We analyze model parametrization in order to obtain an efficient MCMC sampler, and address model selection. The inferential framework described here accommodates any missing-data pattern, and turns out to be useful to analyze statistical relationships among covariates. Finally, we assess the financial
impact of using the covariates, and of the optimal use of the whole available sample when combining data from different mortality experiences.

Original languageEnglish
JournalInsurance: Mathematics and Economics
Early online date14 Jan 2020
DOIs
Publication statusE-pub ahead of print - 14 Jan 2020

Fingerprint

Specifications
Uncertainty

Keywords

  • Mortality
  • Survival models
  • Longevity risk
  • Missing data
  • Mortality models with covariates,
  • MCMC
  • Model selection
  • Bayesian inference

Cite this

@article{b6175e6fbc8b4934b087184a7495bddf,
title = "A hierarchical model for the joint mortality analysis of pension scheme data with missing covariates",
abstract = "A hierarchical model is developed for the joint mortality analysis of pension scheme datasets. The proposed model allows for a rigorous statistical treatment of missing data. While our approach works for any missing data pattern, we are particularly interested in a scenario where some covariates are observed for members of one pension scheme but not the other. Therefore, our approach allows for the joint modelling of data sets which containdifferent information about individual lives. The proposed model generalizes the specification of parametric models when accounting for covariates.We consider parameter uncertainty using Bayesian techniques. We analyze model parametrization in order to obtain an efficient MCMC sampler, and address model selection. The inferential framework described here accommodates any missing-data pattern, and turns out to be useful to analyze statistical relationships among covariates. Finally, we assess the financialimpact of using the covariates, and of the optimal use of the whole available sample when combining data from different mortality experiences.",
keywords = "Mortality, Survival models, Longevity risk, Missing data, Mortality models with covariates,, MCMC, Model selection, Bayesian inference",
author = "Francesco Ungolo and Torsten Kleinow and Macdonald, {Angus Smith}",
year = "2020",
month = "1",
day = "14",
doi = "10.1016/j.insmatheco.2020.01.003",
language = "English",
journal = "Insurance: Mathematics and Economics",
issn = "0167-6687",
publisher = "Elsevier",

}

TY - JOUR

T1 - A hierarchical model for the joint mortality analysis of pension scheme data with missing covariates

AU - Ungolo, Francesco

AU - Kleinow, Torsten

AU - Macdonald, Angus Smith

PY - 2020/1/14

Y1 - 2020/1/14

N2 - A hierarchical model is developed for the joint mortality analysis of pension scheme datasets. The proposed model allows for a rigorous statistical treatment of missing data. While our approach works for any missing data pattern, we are particularly interested in a scenario where some covariates are observed for members of one pension scheme but not the other. Therefore, our approach allows for the joint modelling of data sets which containdifferent information about individual lives. The proposed model generalizes the specification of parametric models when accounting for covariates.We consider parameter uncertainty using Bayesian techniques. We analyze model parametrization in order to obtain an efficient MCMC sampler, and address model selection. The inferential framework described here accommodates any missing-data pattern, and turns out to be useful to analyze statistical relationships among covariates. Finally, we assess the financialimpact of using the covariates, and of the optimal use of the whole available sample when combining data from different mortality experiences.

AB - A hierarchical model is developed for the joint mortality analysis of pension scheme datasets. The proposed model allows for a rigorous statistical treatment of missing data. While our approach works for any missing data pattern, we are particularly interested in a scenario where some covariates are observed for members of one pension scheme but not the other. Therefore, our approach allows for the joint modelling of data sets which containdifferent information about individual lives. The proposed model generalizes the specification of parametric models when accounting for covariates.We consider parameter uncertainty using Bayesian techniques. We analyze model parametrization in order to obtain an efficient MCMC sampler, and address model selection. The inferential framework described here accommodates any missing-data pattern, and turns out to be useful to analyze statistical relationships among covariates. Finally, we assess the financialimpact of using the covariates, and of the optimal use of the whole available sample when combining data from different mortality experiences.

KW - Mortality

KW - Survival models

KW - Longevity risk

KW - Missing data

KW - Mortality models with covariates,

KW - MCMC

KW - Model selection

KW - Bayesian inference

U2 - 10.1016/j.insmatheco.2020.01.003

DO - 10.1016/j.insmatheco.2020.01.003

M3 - Article

JO - Insurance: Mathematics and Economics

JF - Insurance: Mathematics and Economics

SN - 0167-6687

ER -