The effects of parameter uncertainty in the extreme event frequency-severity model

BACK
Jakub M. Borowicz~James P. Norman, United-Kingdom
Download

Summary:
In this paper we analyse parameter uncertainty in the extreme event frequency-severity model. Three methods are compared: classical asymptotic statistics, bootstrapping and a Bayesian approach. The Bayesian method makes use of Markov Chain Monte Carlo techniques, which have recently become more feasible through the use of specialised software and fast computers. These approaches are demonstrated using a data set of large Danish fire losses, previously analysed by several authors. The effects of parameter uncertainty on capital requirements and the price of an Excess of Loss reinsurance contract are investigated in a Dynamic Financial Analysis (DFA) framework. Our results show that, in this case, parameter uncertainty has a significant effect and should not be ignored in DFA models.
 
Date: 30 May - Time: 8:30 to 10:00 - Room: 241
Theme: 3.B. High severity risks and insurability