Optimal investment policy and dividend payment strategy in an insurance company (Q990379)

The article is devoted to dividend optimization problem for an insurance company in the classical Cramér-Lundberg setting. The firm is allowed to control the stream of dividend pay-outs and to invest a part of surplus in a Black-Scholes financial market, and the objective is to find a strategy maximizing the cumulative expected discounted pay-outs until the bankruptcy. The constraint put by the authors on the admissible strategy is that the firm is not allowed to short-sell the stocks and to borrow money to buy stocks (they claim that technically an unconstrained problem is simpler, but unfortunately give no argument for this). The main results of the paper are the following. The authors obtain the optimal value function as the smallest viscosity solution of the associated Hamilton-Jacoby-Bellman equation, and they prove a verification theorem that allows them to recognize the optimal value function among the many viscosity solutions of the associated HJB equation. Under an additional assumption that the claim size distribution function has a bounded density, they show that the optimal value function is twice continuously differentiable except possibly for some points. Then they find the optimal value function for small surpluses and prove that the optimal strategy is stationary, that is, the proportions of surplus invested in the risky asset and paid out as dividends at any time depends only on the current surplus; they moreover prove that the optimal dividend payment policy has a band structure. In particular, the optimal dividend policy for large surpluses is a barrier policy: to pay out immediately the surplus exceeding a certain level as dividends. The authors also obtain the best barrier policy and give examples both where it is optimal and where it is not.