Abstract

BRAVO, Jorge M.  and  AYUSO, Mercedes. Mortality and life expectancy forecasts using Bayesian model combinations: An application to the Portuguese population. RISTI [online]. 2020, n.40, pp.128-145.  Epub Dec 30, 2020. ISSN 1646-9895.  https://doi.org/10.17013/risti.40.128-145.

In this paper we forecast age-specific mortality rates and life expectancy measures using an adaptive Bayesian Model Ensemble of heterogeneous models. The procedure involves both the selection of the model confidence set and the determination of optimal weights. Model-averaged Bayesian credible prediction intervals are derived accounting for both the uncertainty arising from model error and parameter uncertainty. The models are calibrated to Portuguese population data in the period 1960-2018 and age range 60-125 and forecasted to 2050. The results anticipate a continuous increase in both period and cohort life expectancy at adult ages and signal significant implicit taxes and subsidies between generations, which distort consumption, saving and investment decisions and challenge intergenerational fairness.

Keywords : Bayesian Model Ensemble; life expectancy; forecasting methods; stochastic mortality models; longevity risk..

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf )