A debate over whether rating agencies and modeling firms areoverreacting to recent catastrophe events took an odd turn recentlywhen experts warned that companies will play games with catastrophemodels to maintain current ratings.

During the World Insurance Forum here late last month, the“overreaction” question resurfaced frequently, as executivesdiscussed the lingering impact of Hurricane Katrina and otherwindstorms.

On the one hand, insurer and reinsurer executives areanticipating increased prices and higher demand for catastrophereinsurance protection that new versions of cat models will drivelater this year. On the other hand, even modeling firm executivesare worried about increasing rating agency requirements that hingeon loss estimates generated by updated cat models.

“I would hate to see a company being downgraded or havingproblems simply because of a modeling change,” said RichardClinton, president of Eqecat, an Oakland, Calif.-based modelingfirm.

Mr. Clinton and other modeling executives said they would likerating agency analysts to have a better understanding of how modelswork, because they are very different. “They don't produce the samenumbers, and the rating agencies don't take this into account,” hesaid.

Capital charges used in rating agency models depend oncompany-supplied estimates of probable maximum losses, such as100-year windstorms and 250-year earthquakes. “Whatever model youuse” to estimate those losses “is fine with them,” Mr. Clintonreported.

“This creates a moral problem. There is good potential forgaming these models,” he said, noting that insurers could seek touse the model producing the lowest loss estimates for the purposesof a rating, while using other models for underwriting and pricingdecisions.

While new versions of catastrophe models are not scheduled to bereleased until late spring, rating agencies began adjusting theircapital requirements for catastrophic events even before lastyear's storms.

Steven Dreyer, managing director of Standard & Poor's in NewYork, said S&P changed the way it looked at catastrophe risk inresponse to the storms of 2004.

“We recognized that our previous standard didn't deal withaggregate risk very well, so we moved from an event-based approachto an aggregate approach,” Mr. Dreyer explained. In June 2005,S&P replaced a capital charge for reinsurers based on aone-in-100-year event standard with a one-in-250-year aggregatestandard. (After the 2005 storms, S&P said it would also applythe new standard to primary insurers.)

“Our expectation, going forward, is that the modeling agenciesthemselves are adjusting their factors…The net result will be amore onerous capital requirement,” Mr. Dreyer said, noting thatwhile S&P's one-in-250-year standard won't be changed,estimates of aggregate losses occurring once every 250 years willincrease as modelers release new model versions.

Donald Kramer, CEO and president of Ariel Re, believes that byapplying new stress tests to PMLs that will also jump as modelersrework their tools, rating agencies are “double-counting” or“compounding” the impact of catastrophes in their formulas.

“Steve Dreyer said it himself,” Mr. Kramer noted, questioningthe reactions of both the rating agencies and the modeling firms torecent events.

“Is the perception of risk so great that they think a one-in-250is going to happen every year?” he asked. Although he happily notedan opportunity for an underwriter to make money if the perceptionof risk is greater than the actual incidence of loss, he suggestedthat both rating agencies and modelers are overreacting.

“Weather has changed–is changing,” he said. “There will be atime when the Gulfstream shuts down–when Bermuda will be freezing.The question is, 'Is this cyclical? Is Katrina a once-a-year event?Are these events we can expect in 2006?'”

“No company became insolvent as a result of the hurricanes. Theymight have been wounded [or] diminished. Yet companies were forcedout of business by the elimination of ratings,” he added.

His company, Ariel Re, bought the infrastructure of Rosemont Re,which was forced out of business when its ratings were cut. Morerecently, PXRE and Quanta, facing the ratings knife, said theirbusinesses could be sold. Still, Mr. Dreyer noted there weren't alot of downgrades from the 2005 storms, attributing the result tothe revised stress test which put actual losses within S&P'sexpectations.

Mr. Clinton, who refused to be coaxed to describe rating agencyactions as “double-counting,” did speculate that changes inmodeling assumptions and rating agency formulas could, in somecases, produce 50-to-100 percent hikes in capital requirements.“That's going to have a dramatic impact on the industry–on theability to provide capacity to the market,” he said.

Karen Clark, president and CEO of AIR Worldwide in Boston, notedthat “gaming” of models extends to pricing. “There is this quaintnotion that reinsurance pricing is based on the [cat] models, butthe market sets the prices, not the models,” she said.

She noted, for example, that while many industry experts agreeAIR's model has “the most advanced science” for estimating Europeanwindstorm losses, “many prefer other models because the indicatedpricing is cheaper.”

Responding to Mr. Kramer's comments about the gap betweenperception and actual risk, she said, “I don't think [theperception] is high enough. I don't think people have internalizedthe numbers that we [the modelers] have been talking about.”

Over time, she said, “the numbers always go up,” noting that in10 years, loss dollars will be twice as high as they are now justbecause of increasing property values. “One-hundred-billion-dollarevents are going to cost $200 billion,” she added.

Referring to Mr. Kramer's questions about whether Katrina-likeevents are now expected to occur annually, she referred to thescience of cool and warm cycles underlying hurricane activity. “Wecan be in a period now [where there is] a clustering of events. Youhave to be prepared for that.”

She added that “we ought to take these 250- and 100-year returnperiods out of the language,” noting that such terms give a falsesense that an event will only happen once every 250 years. Instead,it means there's a 0.4 percent probability that it will happen thisyear, while a 100-year loss has a 1 percent probability, she noted.“Over 10 years, there's a 10 percent probability that a 100-yearloss will take place.”

Using “return period” terms, RenaissanceRe CEO Neill Currie saidhe empathized with rating agencies and modelers. “When a ratingagency receives something [from an insurer] that says this is my1-in-250, and the actual loss is four times that [amount], how inthe world can a rating agency give someone with that kind of datacredibility?” he asked.

AIR, in an analysis published in November 2005, noted that partof the reason for such discrepancies is the fact that inaccuratedata was being put into catastrophe models. AIR's analysis found,for example, that for 90 percent of commercial properties, insurersrecorded replacement values much lower than values that would beset by an engineering analysis, with some values underestimated byas much as 70 percent.

“You do not want to be at the mercy of this bad data,” Ms. Clarksaid, advising that until data quality is improved, insurers needto benchmark their company loss estimates against industry lossestimates.

While modelers admit to some technical deficiencies in theirmodels, Mr. Currie, Mr. Clark and others said the failure ofinsurers and reinsurers to use models appropriately to makeintelligent underwriting decisions contributed to the surpriseslast year.

“If you are writing risk-excess [reinsurance] and you don't havean occurrence limit for catastrophe, I don't think that's AIR'sfault,” Mr. Currie said. “Underwriting without tools doesn't work,and underwriters without experience relying just on tools doesn'twork, either.”

“There is a lot of na?ve use of the models out there,” agreedEqecat's Mr. Clinton. “I don't know how many times I've heard,'Your model is wrong because it's not matching the market price.'That's not the intention of models,” he said, explaining that theintent is to provide the best science and engineering. “You have todetermine how you're going to price around that along withuncertainty.”

Hemant Shah, president and CEO of Newark, Calif.-based RMS, saidmodeling should be viewed as nothing more than a systematic way oforganizing knowledge. “That actually makes the challenge ofunderwriting that much greater,” he said. “You not only have tounderwrite the account. You need to underwrite the effectiveness ofthat model in interpreting the risks of that account.”

Ms. Clark discussed the need for insurers to use model resultsbased on long-term historical hurricane experience and short-termviews for different types of decisions. In particular, she arguedit would be unlikely primary insurers could get regulatory approvalfor price hikes based on short-term analyses, which will beprovided for the first time in this year's updates. But they coulduse the short-term view to decide how much reinsurance to buy, sheadded.