Help

BI’s Article search uses Boolean search capabilities. If you are not familiar with these principles, here are some quick tips.

To search specifically for more than one word, put the search term in quotation marks. For example, “workers compensation”. This will limit your search to that combination of words.

To search for a combination of terms, use quotations and the & symbol. For example, “hurricane” & “loss”.

Login Register Subscribe

COMMENTARY: Cat modeling makes it hard to differentiate good risk management

Reprints

As underwriters become increasingly reliant on catastrophe modeling, it is becoming harder for risk managers who deploy good property loss control to differentiate their risks from those who do not.

In fact, two risk managers speaking during the recent Risk & Insurance Management Society Inc. conference in Los Angeles became downright angry while addressing the application of Risk Management Solutions Version 11 model for U.S. windstorms, which virtually doubled their organizations' property catastrophe exposures overnight.

“It's confusing to me as a risk manager how I can have a big property in Orlando that one day the model says is going to react like this to hurricanes, and the next day it's going to react totally differently because the winds don't decay at the same speed they thought they decayed,” said Steve Wilder, Burbank, Calif.-based vice president, risk management for The Walt Disney Co. “The models tend to be very generic, so unless you have a lot of information specific to your risks, the models are going to assume either average or worst-case (scenario), so they're not really going to be representative of the risk, and it's going to be really hard for those companies that manage risk proactively and efficiently to differentiate themselves from those that don't.”

Mr. Wilder shared an encounter with a Bermuda underwriter after Hurricane Katrina who said she no longer could participate in Disney's property program “because the models told them that the losses would be devastating. I asked her, "Have you ever been to Disney World?' She said no. "Do you have any engineers that you could send there?'” Her answer was “No,” he said. After that meeting, Mr. Wilder hired RMS, a computer modeling firm based in Menlo Park, Calif., to build a custom model for Disney, which “helped greatly,” he said.

While at RIMS this year, Mr. Wilder met with several property insurers that gave him widely varied estimates of Disney's property catastrophe exposures based on their computer models.

“These numbers are just all over the map,” he said. “I think it's incumbent on us as risk managers to try to differentiate our risks whenever we can. But in the modeling world, it becomes harder and harder to do.”

Scott Clark, risk and benefits officer for Miami-Dade County Public Schools, related a similar experience.

“On an $8 billion schedule, my (probable maximum loss) for a 250-year storm went from $800 million to $1.8 billion,” he said. “The risk didn't change. In fact, it's continued to improve.”

Mr. Clark hired an independent engineering firm to evaluate the school district's property risks, which came to $750 million. When he presented this estimate to RMS, he was told that the engineering firm had used incorrect International Organization for Standardization codes. But even after the correct codes were inserted, “it went back to an $800 million PML for a 250-year storm,” he said.

“Question everything,” Mr. Clark said.