Three Questions: S. Ming Lee, AIR Worldwide Corp.

Modeling has become an increasingly important—and sophisticated—aspect of the insurance and reinsurance process. S. Ming Lee, president and chief executive officer of Boston-based AIR Worldwide Corp., discussed trends in catastrophe modeling Monday with Business Insurance Special Projects Editor Rodd Zolkos.

Q: In a mild hurricane year like 2009 is shaping up to be, do you think insurers sometimes forget some catastrophe possibilities?

I think people actually have a fairly short memory. These events can happen, they do happen and they can happen at any time. And the models show that, and it's important to look at the models' results and understand the results and understand the possibilities.

It's our role to do a good job with the science and make sure that we've got all the latest science and the engineering, and the model results will then show what the relative risk is.

Q: What's the state of the art of catastrophe modeling?

The models started over 20 years ago. And they've evolved quite a lot, especially those models which have had many events over the historical period. The windstorm models, of course, got refined with lots of new claims information and loss experience.

Fortunately, earthquake—the physical phenomena of earthquake—is similar in different parts of the world. For example, in Turkey the major fault there is very similar to the San Andreas Fault in California.

So when earthquakes occur throughout the world, we learn from them, and we can apply what we learn to other parts of the world, and that helps to improve the earthquake models.

Of course, earthquakes don't happen as frequently as the windstorms. (But) you can continue to learn from all of them. Also, it's not only just the hazard. There is continuing research on the vulnerability of structures. There is continuing research on ground motion, and this new research is applied to the revised earthquake models. So the earthquake models can also improve, even though we don't see as many large earthquakes.

Q: Following Katrina there was considerable discussion of the importance of data quality. Is there still an emphasis on improving data?

We've undertaken several major initiatives. One of them is the development of our TruExposure solution set for improving the quality of exposure data, and this will be available toward the end of the year, for reinsurers as well as for insurers, to do a number of things.

One is to validate the exposure data for reasonableness, for completeness. Second is to benchmark their data against our industry averages, and then third is to augment the poor data with better data from specific property databases.

Basically the database for augmentation is derived from many sources, the primary sources being the specific property information database developed by (Insurance Services Office) over the years from physical inspection of buildings.