COMMENTARY: Storm-risk models impact rates

Bradenton Herald--Feb. 28, 2008

By ALDEN WEICHEL
Special to the Herald

Insuring hurricane losses in Florida is a high-risk business. Most of the insured property – worth nearly $2 trillion – is vulnerable to storm losses since Florida has more major hurricanes than any other place in the world.

The primary job of state regulators is to make sure that companies writing policies here have the financial resources to pay claims. Increasingly, companies and state regulators use “catastrophe models” or “risk models” to determine just how much a company could lose and how much cash it needs to have on hand to pay its claims after a hurricane, as well as how much it would need to continue to write new policies.

Risk models have an enormous impact on the cost and availability of property insurance in Florida. Their complex mathematical formulas are confusing and, at times, contradictory. For instance, a risk model formulated by state officials last year projected that insured losses from a major storm would be twice the amount projected by independent risk models.

Despite their flaws, the models are the best tools we have to determine how best to solve the vexing problem of making insurance more affordable and available in Florida.

In fact, the problem may not be with the models’ formulas but with the way that insurers, regulators and financial-rating agencies react – or overreact – to their projections. Balancing insurers’ need for financial solvency with consumers’ need for affordable insurance protection is not easy when the risk of a major storm is so unpredictable and the potential loss so devastating.

Catastrophe models use detailed computer models and thousands of model simulations. They provide an independent, objective assessment of the catastrophe loss potential facing any insurance company. Since all insurance companies buy reinsurance to enable them to pay claims resulting from catastrophes such as large hurricanes, the reinsurers’ catastrophe models and interpretations of these dictate the insurance company’s cost of the reinsurance. This cost is passed on to you as a homeowners policyholder.

The active hurricane seasons of 2004 and 2005, as well as the lack of such violent storms in 2006 and 2007, has prompted a number of questions in the insurance industry as to whether hurricanes will follow a fairly predictable cyclical weather pattern or if storm experience will now be less predictable and more severe.

Catastrophe models do not predict the future. But they are analytical tools that support decision-making.