二维分散逼近模型的最优比例再保险和分红策略.pdfVIP

℄ b3  1 b3   HJB  HJB   : , , , HJB i Æ OPTIMAL PROPORTIONAL REINSURANCE AND DIVIDEND CONTROL FOR A TWO-DIMENSIONAL DIFFUSION APPROXIMATION ABSTRACT This paper consider a risk model which can choose a risk control and dividend control. The objective is to find the optimal policy which max imizes the expected total discounted dividend payout until the time of bankruptcy. We model the dynamics of the corporate wealth as a two dimensional diffusion approximation with the diffusion and drift coefficients being affine function of the risk control variable. We show that there exists an optimal level b3 such that the optimal dividend policy is to keep the company’s wealth below b3 and pay out as dividends to the shareholders all the amount in excess of this level. On the other hand, the optimal risk control policy depends on the ratio of reinsurance safety loading to insur ance’s. When this ratio is big, the optimal policy is always to undertake the maximal risk. When this ratio is small, the risk depends on the current amount of the wealth. The risk as a function of the wealth is monotone and increasing, reaching the maximum at some point smaller than b . 3 This paper includes six chapters.