Stochastic control in insurance

The objective of this research project is to apply stochastic control methods in problems related to insurance, such as problems of finding optimal reinsurance strategies, premium strategies or investment strategies. For the latter we consider both problems in damage and life insurance, while the two former is only concerned with damage insurance. As criteria for optimality we mainly use expected pay-out of dividends, but other criteria such as minimization of ruin probabilities are also considered. The work is mainly based on the diffusion model but there are also found results for the classical Cramer-Lundberg model. Cooperation with Michael Taksar,University of Missouri,USA and Elena Vigna, University of Turin, Italy.