Parsimonious parameterization of age-period-cohort models by Bayesian shrinkage. (English) Zbl 1390.62226

Summary: Age-period-cohort models used in life and general insurance can be over-parameterized, and actuaries have used several methods to avoid this, such as cubic splines. Regularization is a statistical approach for avoiding over-parameterization, and it can reduce estimation and predictive variances compared to MLE. In Markov Chain Monte Carlo (MCMC) estimation, regularization is accomplished by the use of mean-zero priors, and the degree of parsimony can be optimized by numerically efficient out-of-sample cross-validation. This provides a consistent framework for comparing a variety of regularized MCMC models, such as those built with cubic splines, linear splines (as ours is), and the limiting case of non-regularized estimation. We apply this to the multiple-trend model of A. Hunt and D. Blake [“A general procedure for constructing mortality models”, North Am. Actuar. J. 18, No. 1, 116–138 (2014; doi:10.1080/10920277.2013.852963)].


62P05 Applications of statistics to actuarial sciences and financial mathematics
62J07 Ridge regression; shrinkage estimators (Lasso)
91B30 Risk theory, insurance (MSC2010)
Full Text: DOI Link


[1] Antonio, K.; Bardoutsos, A.; Ouburg, W., Bayesian poisson log-bilinear models for mortality projections with multiple populations, European Actuarial Journal, 5, 245-281, (2015) · Zbl 1329.91111
[2] Barnett, G.; Zehnwirth, B., Best estimates for reserves, PCAS, 87, 245-303, (2000)
[3] Blei, D. M., (2015)
[4] Cairns, A.; Blake, D.; Dowd, K.; Coughlan, G.; Epstein, D.; Ong, A.; Balevich, I., A quantitative comparison of stochastic mortality models using data from England and Wales and the United States, North American Actuarial Journal, 13, 1-35, (2009)
[5] Calderhead, B.; Radde, N., Hamiltonian Monte Carlo methods for efficient parameter estimation in steady state dynamical systems, BMC Bioinformatics, 15, 253, (2014)
[6] Carlson, E., The Lucky Few, (2008), Netherlands: Springer, Netherlands
[7] Casella, G.; George, E. I., Explaining the Gibbs sampler, The American Statistician, 46, 167-174, (1992)
[8] Chung, Fung, M.; Peters, G.; Shevchenko, P., (2016)
[9] Efron, B.; Morris, C., Data analysis using Stein’s estimator and its generalizations, Journal of the American Statistical Association, 70, 311-319, (1975) · Zbl 0319.62018
[10] Fienberg, S. E.; Mason, W. M., Identification and estimation of age-period-cohort models in the analysis of discrete archival data, Sociological Methodology, 10, 1-67, (1978)
[11] Frost, W. H., The age selection of mortality from tuberculosis in successive decades, American Journal of Hygiene, 30, 91-96, (1939)
[12] Gelfand, A. E.; Gilks, W. R.; Richardson, S.; Spiegelhalter, D. J., Markov Chain Monte Carlo in Practice, Model determination using sampling-based methods, 145-162, (1996), London: Chapman and Hall, London
[13] Gelfand, A. E.; Dey, D. K.; Chang, H., (1992)
[14] Greenberg, B. G.; Wright, J. J.; Sheps, C. G., A technique for analyzing some factors affecting the incidence of syphilis, Journal of the American Statistical Association, 45, 373-399, (1950)
[15] Hachemeister, C. A.; Stanard, J. N., IBNR claims count estimation with static lag functions, (1975), ASTIN Colloquium: ASTIN Colloquium, Portimão, Portugal
[16] Hastie, T.; Tibshirani, R.; Wainwright, M., Statistical Learning with Sparsity, (2015), Boca Raton, FL: CRC Press, Boca Raton, FL · Zbl 1319.68003
[17] Hoerl, A.; Kennard, R., Ridge regression: Biased estimation for nonorthogonal problems, Technometrics, 12, 55-67, (1970) · Zbl 0202.17205
[18] Hoffman, M. D.; Gelman, A., Hamiltonian Monte Carlo methods for efficient parameter estimation in steady state dynamical systems, Journal of Machine Learning Research, 15, 1351-1381, (2014)
[19] Hunt, A.; Blake, D., A general procedure for constructing mortality models, North American Actuarial Journal, 18, 116-138, (2014)
[20] Elementary Calculus: An Infinitesimal Approach, (2000), Prindle, Weber & Schmidt
[21] Klugman, S. A.; Panjer, H. H.; Willmot, G. E., Loss models: From Data to Decisions, (2008), New York: Wiley, New York · Zbl 1159.62070
[22] Lee, R. D.; Carter, L. R., Modeling and forecasting U.S. mortality, Journal of the American Statistical Associationl, 87, 659-675, (1992) · Zbl 1351.62186
[23] Mcdonald, J., Some generalized functions for the size distribution of income, Econometrica, 52, 647-663, (1984) · Zbl 0557.62098
[24] Ntzoufras, I., (2010)
[25] Van Ravenzwaaij, D.; Cassey, P.; Brown, S., (2016)
[26] Renshaw, A. E.; Haberman, S., A cohort-based extension to the Lee-Carter model for mortality reduction factors, Insurance: Mathematics and Economics, 38, 556-570, (2006) · Zbl 1168.91418
[27] Ryder, N. B., The cohort as a concept in the study of social change, American Sociological Review, 30, 843-861, (1965)
[28] Taylor, G., Separation of inflation and other effects from the distribution of non-life insurance claims delays, ASTIN Bulletin, 9, 217-230, (1977)
[29] Vehtari, A.; Gelman, A.; Gabry, J., Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC, Statistics and Computing, 27, 1413-1432, (2017) · Zbl 06737719
[30] Venter, G. G., Transformed beta and gamma distributions and aggregate losses, Proceedings of the Casualty Actuarial Society, LXX, 156-193, (1983)
[31] Verbeek, H. G., An approach to the analysis of claims experience in excess of loss reinsurance, ASTIN Bulletin, 6, 195-202, (1972)
[32] (2016)
[33] Xu, Y.; Sherris, M.; Ziveyi, J., The application of affine processes in multi-cohort mortality model, (2015), University of New South Wales Business School Research
[34] Ye, J., On measuring and correcting the effects of data mining and model selection, Journal of the American Statistical Association, 93, 120-131, (1998) · Zbl 0920.62056
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.