We are pleased to introduce a new series of articles provided by Associate Members of the Oasis community. The articles will cover market industry topics packed with useful information. Enjoy the read and please do get in touch with us, if you have any comments, questions or queries: info@oasislmf.org.
The first article focuses on the Challenges in Catastrophe Modelling by Dr. Nigel Winspear, CATRISX SERVICES Ltd.
Introduction
Catastrophe (CAT) models have added enormous amounts of value over the past three decades, by standardising the simulation of extreme CAT events and their risk metrics across the (re)insurance industry. They also routinely provide a sound basis for CAT risk accumulation, capital management and technical pricing across the industry; and can take a lot of credit for reducing the number of insurer insolvencies in the wake of large, unexpected CAT events. Models can and do face limitations, however, as we all do and as the modellers of the coronavirus pandemic are currently finding. For starters, the real world is a very complex place and resists efforts to reduce it to the confines of a model with relatively few, incompletely understood, often complexly interacting parameters. Nature also throws complete surprises at us, as happened in March 2011 with the unexpected magnitude of the subduction megathrust rupture of the Japan trench in the Tohoku earthquake, after which modellers played catch up. CAT modelling is not a finished art, there are still key areas needing further attention from model developers, such as the five challenges presented here. The purpose of this paper is to galvanise model vendors and model evaluation specialists into action to jointly meet these challenges.