WASHINGTON--While the commission charged with evaluating hurricane loss models for the state of Florida approved one model, it rejected another and faces criticism from that company that it is not adapting to newer methods of predicting catastrophic losses.

Newark, California-based Risk Management Solutions, Inc., was set to have its v6.0 RMS U.S. Hurricane Model reviewed by the commission on Thursday. The company has since pulled the model back for adjustments after being told by a professional team from the Florida Commission on Hurricane Loss Projection Methodology (FCHLPM) that the model would not meet the commission's standards.

The commission, established in 1995, serves as an independent evaluator of computer loss projection models and other actuarial methods for projecting losses in the event of a hurricane.

The commission approved the model of another company, Oakland, Calif.-based EQECAT. It also certified Boston-based AIR Worldwide Corporation's 2007 U.S. Hurricane Model.

Models are certified on an annual basis by the FCHLPM for use by insurers for residential rate filing. AIR said its model has met the commission's standards for 11 consecutive years.

RMS said it had been in discussions with the commission over a "new method" to gauge hurricane risk based on the opinion drawn from the expertise of the hurricane research community rather than a historical average that does not factor in periods of higher or lower hurricane frequency.

The company added that the 6.0 model used this approach in addition to the traditional, historical average, and was disappointed when a professional team sent by the commission to examine the model was unable to verify that it would meet all of the commission's criteria.

"We continue to recommend that insurance and reinsurance companies use the standard version of the RMS model as the basis for assessing and managing risks," said Mitch Sattler, vice president of public policy at RMS. "This incorporates forward-looking estimates of medium-term hurricane activity over the next five years. Once certified by the FCHLPM, the version based on the long-term view should be used for compliance within the Florida regulatory system."

Mr. Sattler noted that, "We communicated our disappointment to the Florida Commission regarding the continued disconnect between the current regulatory standards and science, and the resulting implications for misunderstanding hurricane risk in Florida."

He said RMS is "committed to working with the FCHLPM to change what are now outdated standards that limit Cat models to use of the long-term historical average."

A spokesperson for FCHLPM said RMS was within its rights to pull back the 6.0 model, adding that the model is now scheduled for review on June 20.

Mr. Sattler explained that the long-term historical average significantly underestimates the level of hurricane hazard along the U.S. coast, "and there is a consensus among expert hurricane researchers that we will continue to experience elevated frequency for at least the next 10 years. The current standards make it more difficult for insurers and their policy-holders to understand, manage, and reduce hurricane risk effectively."

EQECAT's North Atlantic Hurricane model, first released in 1995, was designed to project potential losses of a storm striking the mainland United States, the Bahamas, Bermuda and other Caribbean islands.

"The recertification success of EQECAT's model is attributable largely to our ongoing dedication to making enhancements to meet the specific needs of the insurance and reinsurance community," said Rick Clinton, president of EQECAT. "When we enhance our model, we focus specifically upon the consistent representation of risk across the North Atlantic basin; the appropriate per-occurrence loss exceedance curve based on climatologically reasonable hurricane tracks and parameter histories; and the appropriate spatial correlation of loss."