Cat Models An Issue For NAIC Confab

An insurers group said when the nation's insurance regulators meet this week, it will urge them to allow insurers to use the best natural catastrophe loss models available to manage risks.

The American Insurance Association, Washington, will make its case during a hearing Friday that will be held during the fall meeting of the National Association of Insurance Commissioners in Washington.

Eric Goldberg, AIA associate general counsel, said in a statement that catastrophe models are very important for a stable insurance market. Cat models use computers to estimate potential loss to property due to a catastrophic event.

Some models have come under criticism from consumer advocates, who contend that their projections on the amount of future storm activity are incorrect and serve as an improper basis for raising rates.

Other issues to be discussed, according to AIA, include collateral requirements for unlicensed reinsurers and climate change. AIA said it wants a dialogue on climate change but opposes proposed interrogatories relating to the issue in annual financial statements, said David Snyder, AIA vice president and assistant general counsel.

Collateral requirements for unlicensed reinsurers will also receive discussion, and the AIA supports the current collateralization system that provides reinsurers with the option to become licensed in the U.S. or post 100 percent collateral and not submit to U.S. licensing requirements, according to Steven Bennett, AIA general counsel.

The Property Casualty Insurers Association of America, Des Plaines, Ill., maintained that cat risk models are “the most sophisticated tools available for the assessment of extreme risk exposure,” said David Kodama, director-policy analysis and policy development and research for PCI.

Without the ability to price according to the full extent of risk, regulatory constraints can increase the potential for an insurer to become insolvent, he added.

A one-size-fits-all set of assumptions limits the potential use of multiple models to gain greater information, Mr. Kodama noted.