Clarifying Cat Models

With Florida and Louisiana regulators declining to certify the use of five-year models in the underwriting process, catastrophe modelers must be feeling a little like Albert Einstein who once observed of the disconnect between science and the world beyond the lab: “Science is a wonderful thing if one does not have to earn one’s living by it.† Of course catastrophe models have come under a lot of scrutiny since the record hurricane loss seasons of 2004 and 2005, and it’s important to recognize that they are just one of the many tools that help insurers, reinsurers and risk managers more accurately analyze, write and price for catastrophe risk. The output from a given catastrophe model is also only as good as the data embedded in it and subject to the different variables of an individual insurer’s book of business. That said, today’s models are increasingly sophisticated and constantly being fine-tuned to incorporate the latest technologies and data. Models have also been developed for a wide range of catastrophic risks beyond hurricanes, including earthquakes, tornadoes, floods and terrorism. The use of computer technology in the underwriting process is not new, and in the case of catastrophe models there’s no doubt that with this tool, underwriters are better placed to more accurately analyze and evaluate catastrophe risk going forward. But given recent concerns, clearly there is more to be done in better explaining how the models actually work.  Check out  I.I.I.’s catastrophe modeling update.

Leave a Reply

Your email address will not be published. Required fields are marked *