Catastrophe models will continue to be exposed by events likeSuperstorm Sandy because they can fail to contemplate all possiblesources of loss, says Lockton Cos.

A report by Jeff Tennis, manager of catastrophe analytics atLockton, says model deficiencies are visible "particularly forareas with a limited historical record," but he adds it "would benaïve to fully discredit" these loss-prediction tools.

Tennis suggests simulations using the hurricane model by RiskManagement Solutions (RMS) were not lacking in that the firmmodeled 2,300 storms in the area where Sandy hit and 62 stormsproduced storm surge of 16-feet or greater—the surge height in someareas during Sandy, such as on Staten Island, N.Y.

However, Tennis says, the simulated storms to produce such asurge in the model were a Category 3 or greater. Sandy was aCategory 1, and made landfall as a tropical storm.

"RMS claims its model represented Sandy's storm surge reasonablywell, but has provided few details supporting this claim," Tennissays. 

Additionally, inland flooding in not captured by the RMS model,Tennis adds.

The models also miss the mark on indirect losses such thoserelated to business interruption, which, in an area likeNew York, is a big driver of loss—maybe more so than directlosses.

The point is not that the models are useless. Tennis says models"are still a useful tool to evaluate 'how bad bad can get.'"

The point is insurance professionals, "must recognize modeldeficiencies and supplement where needed."

Read Tennis' report HERE.

RMS says insured losses from Sandy will be between $20 billionand $25 billion. In a December interviewwith PC360, the modeler saysSandy verified its model's approach to estimatingdamages from wind and from storm surge.

Catastrophe modelers AIR Worldwide and Eqecat also releasedinsured-loss estimates from Sandy. AIR raised its estimate ofinsured losses from Superstorm Sandy to between $16 billion and $22billion after releasing a much lower range very soon after thestorm. Eqecat says insured losses will total $10 billion to $20billion.

Industry executives and analysts have since predicted insuredlosses will at least reach the high end of RMS' estimate, if notexclipse the prediction.