A lattice-based approach for life insurance pricing in a stochastic correlation framework

We propose a new implementation approach in insurance product valuation to capture the stochastic correlation between financial and demographic factors. This is important to accommodate the prevailing situation where the interest rate and mortality intensity move jointly and randomly. A stochastic correlation model is considered where it follows a diffusion process that may assume the form of a bounded Jacobi process or of a transformed modified Ornstein–Uhlenbeck process. Our contributions strengthen the general modelling set up of dependent financial and actuarial risks. We put forward a discrete-time pricing model that supports the valuation of a relatively wide class of insurance products. Specifically, the pricing of contracts, with an embedded surrender option for which no explicit formulae are available, is facilitated with ease. We customise the construction of lattice discretisations that admit a large set of risk processes having appropriate specifications. In particular, the interest rate, mortality and correlation dynamics are discretised via three different binomial lattices that are then assembled to create a trivariate lattice structured with eight branches for each node. Numerical experiments involving some stylised insurance contracts are conducted. Such experiments confirm the accuracy and efficiency of our proposed approach with respect to two benchmarks: the Monte-Carlo simulation method, and the method and results by Devolder et al. (2024).