Simulation of ruin probabilities for risk processes of Markovian type. (English) Zbl 1014.91055

Albrecher, Hansjörg; Kantor, Josef

Monte Carlo Methods Appl. 8, No. 2, 111-127 (2002).

Summary: We consider a generalisation of the classical risk model, where consecutive claims may be dependent according to a Markovian structure represented by a copula function for the joint distribution function of the claims. For various marginal claim size distributions and copula functions ruin probabilities are simulated via Monte Carlo and an importance sampling technique for variance reduction is developed.

MSC:

91B30	Risk theory, insurance (MSC2010)
62P05	Applications of statistics to actuarial sciences and financial mathematics

Keywords:

classical risk model; Markovian structure; copula function; ruin probabilities