G. Pflug, S. Hochrainer, A. Timonina
We consider the problem of optimal design of an insurance contract for extreme events under a mean-risk objective. The assumed loss distribution is estimated from past data, but these estimates are quite unprecise given the rare data.
We extend the problem by considering the minimax approach using an appropriate ambiguity model and show how the optimal solution changes under the ambiguity approach.
The problem is motivated by a practical problem from IIASA, Laxenburg.
Keywords: stochastic optimization, ambiguity, minimax, saddlepoints
Scheduled
ThA1 Insurance
June 2, 2016 9:00 AM
Salón de actos
