Claims News Service, Dec. 30, 12:54 p.m. EST — A report on theperformance of near term hurricane models finds that the modelssignificantly overestimated U.S. losses from Atlantic hurricanesfor the cumulative 2006 through 2008 hurricane seasons.

Karen Clark & Company, a catastrophe risk management firm,analyzed near term models that were introduced in 2006 by the threemajor catastrophe modelers – AIR Worldwide (AIR), EQECAT and RiskManagement Solutions (RMS). AIR initially predicted an overallannualized increase in hurricane losses of 40 percent above thelong term average, but later lowered that figure to 16 percent in2007. EQECAT predicted increases of between 35 and 37 percent, andRMS consistently predicted an overall increase of 40 percent abovethe long term average.

Assuming long term average annual hurricane losses of $10billion for each year, these figures translate into cumulativeinsured losses for 2006 through 2008 of $37.2 billion, $40.8billion, and $42 billion respectively, for the AIR, EQECAT and RMSmodels. The actual cumulative losses were $13.3 billion, far lowerthan the model predictions, and more than 50% below the long termcumulative average of $30 billion.

Catastrophe models were introduced to the insurance industry inthe late 1980s. By utilizing many decades of historical data, themodels gave insurance companies better estimates of what couldhappen and more specifically, the probabilities of losses ofdifferent sizes on specific portfolios of insured properties. Thedestructive 2004 and 2005 hurricane seasons were catalysts forintroducing the near term models. Use of these models by insuranceand reinsurance companies, which are based on short termassessments of the frequencies of hurricanes, was a radicaldeparture from the way in which catastrophe average annual losses(AALs) and probable maximum losses (PMLs) are typicallyderived.

According to the Karen Clark & Company report, in order forinsured losses to reach 40 percent above average for the five yearperiod, in line with the highest model predictions, the next twoyears will have to be similar to 2004, or there will have to beanother Hurricane Katrina.

The report notes that hurricane activity is influenced by manyclimatological factors, many of which are known, but some unknown,by scientists. There are complicated feedback mechanisms in theatmosphere that cannot be quantified precisely even by the mostsophisticated and powerful climate models. The report recommendsthat insurers, reinsurers and regulators evaluate the efficacy ofthe near term hurricane models in light of this uncertainty.

The report, Near Term Hurricane Models – How Have TheyPerformed?, is available at www.karenclarkandco.com.

Interested in more catastrophe news and in-depth articles? Headover to Claims' catastrophe channel for more information.