Abstract
This paper studies the use of simulation techniques to model complex situations in insurance business, where the risk process for old and new business are modeled by compound Poisson processes. The simulation is used to benchmark the solutions of the Hamilton-Jacobi-Bellman equation and to obtain results in the case of distributions, where solutions cannot be obtained analytically. We also explore the problem of estimating the probability of achieving a target capital before ruin.
Additional information
Notes on contributors
M. Kelbert
Mark Kelbert He is a Reader in Statistics at Department of Mathematics, Swansea University. His research interests include queuing theory, diffusions and branching diffusions on manifolds, and mathematical physics.
I. Sazonov
Igor Sazonov He is a research officer at Civil & Computational Engineering Centre, Swansea University. His research interests include numerical modeling, fluid mechanics, acoustics, finite element methods, mesh generation.
A. P. Shore
Adam Shore He obtained BSc(Hons) in Actuarial Science from University of Wales — Swansea in 2002. Since then, he has been doing a PhD research in stochastic processes in the School of Business and Economics, Swansea University. Among his interests are stochastic control problems in insurance and dynamic programming.