Combined impulse control and optimal stopping in insurance and interest rate theory.

Abstract

In this thesis, we consider the problem of portfolio optimization for an insurance company with transactional costs. Our aim is to examine the interplay between insurance and interest rate. We consider a corporation, such as an insurance firm, which pays dividends to shareholders. We assume that at any time t the financial reserves of the insurance company evolve according to a generalized stochastic differential equation. We also consider that these liquid assets of the firm earn interest at a constant rate. We consider that when dividends are paid out, transaction costs are incurred. Due to the presence of transactions costs in the proposed model, the mathematical problem becomes a combined impulse and stochastic control problem. This thesis is an extension of the work by Zhang and Song [69]. Their paper considered dividend control for a financial corporation that also takes reinsurance to reduce risk with surplus earning interest at the constant force p > 0. We will extend their model by incorporating jump diffusions into the market with dividend payout and reinsurance policies. Jump-diffusion models, as compared to their diffusion counterpart, are a more realistic mathematical representation of real-life processes in finance. The extension of Zhang and Song [69] model to the jump case will require us to reduce the analytical part of the problem to Hamilton-Jacobi-Bellman Qausi-Variation Inequalities for combined impulse control in the presence of jump diffusion. This will assist us to find the optimal strategy for the proposed jump diffusion model while keeping the financial corporation in the solvency region. We will then compare our results in the jump-diffusion case to those obtained by Zhang and Song [69] in the no jump case. We will then consider models with stochastic volatility and uncertainty as a means of extending the current theory of modeling insurance reserves.