On the Time to Ruin for Erlang(2) Risk Model in a Markov Environment

In order to measure the increasing complexity and dependent risk of nonlife insurance products and models, a class of the renewal risk processes with non-stationary and stochastic dependence properties are considered in this paper. By introducing an external continuous-time Markov process, the generalized Erlang(2) risk model can rationally characterize the dependent structure, in which the interclaim time, the claim amount and the premium rate are all regulated by the Markov process. The Gerber-Shiu discounted penalty functions (GS functions) are utilized to deal with the ruin probabilities in this model. The defective renewal equations are derived from taking the Laplace transform of the integro-differential equations that the GS functions satisfy. This Markov-modulated Erlang(2) risk model can effectively measure a type of dependent risk.