Modelling Mortality using Multiple Stochastic Latent Factors

Abstract

In this paper we develop a new model for stochastic mortality that considers the possibility of both positive and negative catastrophic mortality shocks. Specifically, we assume that the mortality intensity can be described by an affine function of a finite number of latent factors whose dynamics is represented by affine-jump diffusion processes. The model is then embedded into an affine-jump framework, widely used in the term structure literature, in order to derive closed-form solutions for the survival probability. This framework and model application to the classical Gompertz-Makeham mortality law provides a theoretical foundation for the pricing and hedging of longevity-linked derivatives.