Insurance Industry Relies on Flawed Models to Set Rates ~

In theory, premium rates for all kinds of insurance - auto coverage, life insurance, health insurance, and homeowners insurance - are set by a sophisticated examination of the actuarial risk involved in the value of the property being insured, and the nature of the behavior or activity that creates a hazard to that value. In the case of insurance against hurricane damage, however, the Sarasota Herald-Tribune is reporting this month that insurance company risk-modeling programs are fraught with error, misinformation, and wild guess-work (" Insurers' computer models deeply flawed," by Paige St. John). Complex computer models used by insurance companies, according to the paper, are a "black box" whose contents are kept secret from all prying eyes - with the exception of one obscure Florida agency, the Florida Commission on Hurricane Loss Projection Methodology. That agency knows something about how the models work, but is bound to keep those details confidential as trade secrets, in spite of Florida's open-government "Sunshine" laws. Developed decades ago as advisory tools, the paper says, these top-secret computer models are now relied on as a source of hard, precise numbers used to set premiums, and to purchase reinsurance in case the insurance companies bite off more than they can chew - even though the programs' output is at best a range of possibilities, similar to the cone of probabilities generated by weather forecasters to approximate the predicted landfall of a hurricane. Inaccuracies in the models may come out in the wash over time, as a large number of storms and a large number of insurance companies and homeowner policies are factored in, the paper says. But for a given company handling a given storm, “the final number chosen from the range of estimates ... is almost always wrong.” Worse, according to the report, insurance companies working to set premiums or to influence state officials, given multiple computer models to choose from, have gotten into the habit of choosing the model whose answer they like the best in any given case. "The Herald-Tribune found insurers choosing models or using them in ways that boosted their bottom line, including to argue for rate hikes," said the report. "Filings with Florida regulators show several insurers sought rate increases this year after using catastrophe models that left out loss-reducing details such as roof shape or storm shutters... State Farm this year openly switched to a model that better supported its rate-hike request. After years of presenting other models, State Farm swapped to a seldom-used engineering-based catastrophe model to argue for a 21 percent rate increase on homes fortified against hurricanes." With complexity in the risk analysis has come increasing cost for individuals and for states - along with reduction in coverage and a confusing gauntlet homeowners must run in the event of a catastrophic storm. The RAND Corporation think-tank released a study of the troubled insurance market in October, pointing to, among other vexing policy issues, a pattern of rising premiums and reduced coverage (" Residential Insurance on the U.S. Gulf Coast in the Aftermath of Hurricane Katrina," by James W. Macdonald, Lloyd Dixon, and Laura Zakaras). Says the report, "In areas exposed to the most risk of wind and flood damage, prices for private residential wind insurance have increased dramatically at the same time that access to coverage has declined." And as private insurers have pulled out of the riskiest markets, state wind pools have stepped in to pick up the slack - making the situation, if anything, more complicated for homeowners. "Because homeowners insurers often exclude wind coverage (along with flood coverage)," the report notes, "many homeowners in coastal regions must now purchase three insurance policies to insure the same dwelling: one, underwritten by a traditional insurer, to cover perils such as fire and theft (but not wind and flood); a second, from the state windstorm residual market, to cover damages from wind or hail; and a third, from the NFIP, to cover flooding." Typically, the policies have different deductibles and other key details - and they also require homeowners who suffer a loss to somehow establish whether wind, flood, fire, or some combination of those forces caused their damage. Meanwhile, insurance companies continue to try to predict the future - and, according to the Herald Tribune, to fudge the results in their own financial interest. According to the paper, Florida insurance companies are still relying on storm predictions based on the work of a handful of scientists, some of whom no longer support the conclusions of their collaboration (" Florida insurers rely on dubious storm model," by Paige St. John). After Hurricane Katrina, the paper reports, an insurance industry consulting firm gathered four weather scientists together in Bermuda to ask how many storms would hit the U.S. - and, in particular, Florida - in the next five years. The firm, RMS, used the results to help persuade Florida regulators to allow a sharp escalation of insurance premiums in the state — saying that the data represented “scientific consensus,” reports the Herald-Tribune. “The reality was quite different,” the paper asserts: “Today, two of the four scientists present that day no longer support the hurricane estimates they helped generate. Neither do two other scientists involved in later revisions. One says that monkeys could do as well.” That scientist — University of Colorado professor Roger Pielke — posts about the continuing scientific controversy on his blog (“ RMS Responds to the Sarasota Herald-Tribune”). Comments Pielke, “Along with its peers, RMS is an important company. They do work that potentially helps make the global reinsurance and insurance industry do its work with a closer connection to empirical science. It is precisely because RMS is so important that it merits close attention. Like ratings agencies, RMS and other catastrophe modelers are too important to a range of public outcomes to be left to govern themselves.” After Katrina, the Herald-Tribune says, "The new RMS model called for at least 11 hurricanes to come ashore in the United States by the end of 2010, most of them aimed at Florida." And what happened in reality? "Four hurricanes struck the U.S. None hit the Sunshine State."