Insurers are better prepared to face the hurricane season this year, after the harsh 2004 season forced them to learn how to more effectively interpret the information they get from catastrophe (CAT) modeling systems, according to Risk Management Solutions (RMS), a Newark, Calif., provider of products and services for the management of catastrophe risk. At press time, RMS estimated that losses from Hurricane Dennis were likely to be between $1 and $3 billion.The two million claims produced from the 2004 season were a catalyst that encouraged companies to improve their understanding of the models, much like Hurricane Andrew in 1992 pushed catastrophe modeling into the mainstream, says Kyle Beatty, meteorologist with RMS.
"We felt an obligation to fairly investigate the claims produced from the 2004 season," he says.
Beatty claims the hurricanes last year forced carriers to pay more attention to smaller variables, particular to specific hurricanes. For instance, insurers dealt with a limited supply of materials for repairs, increasing the overall cost of settling claims.
During Hurricane Dennis, they still faced the residual effects of this demand surge. This year, notes Beatty, carriers are better able to incorporate details like the changing prices of rebuilding supplies into the final loss estimates, making them more accurate. The data entered into a model can be as simple as a zip code, points out Alberto Leal, actuarial director at Fireman's Fund, Novato, Calif.
The more sophisticated models now being used can include intricate data, such as which portion of a building is masonry, the shape of the masonry, and where it is located in the building, he says. "A building's specific vulnerabilities and characteristics can change loss estimates significantly," notes Leal.
Interpretation key
"It appears you can't simply apply an average differential to every state, or even every county," Leal adds. "The better you are at interpreting this information, the better off your company is, the better off the industry is. There's a lot of sophistication currently, and it's growing tremendously."
However, insurers must decide how much data is practical, or it can get overwhelming, stresses Leal. Third-party vendors can take some of the sting out of the labor-intensive nature of slicing and dicing the data by providing information about a potential investment's characteristics, but the resource is costly, he says.
At press time, several insurance companies were using the models to help better respond to Hurricane Dennis, and were using the models to monitor Hurricane Emily. The information garnered from the models allows companies to report potential losses days before the storm hits land, and send agents and adjusters out early to affected areas.
"We are running the variables to provide our clients with information about the size of loss," says Julie Serakos, senior vice president of London-based Willis, an RMS customer. "Using that information, the carriers will figure out how to respond."
Whereas most carriers are using CAT technology to gear up for weather-related catastrophes, some are leveraging CAT modeling systems to help evaluate risks and diversify, and therefore charge lower premiums.
And with the advances in CAT modeling technology not withstanding, Beatty says the most important steps have been an increase in the ability to understand and use the information from the models.
"The industry in general is better prepared to use these products," says Beatty.
