EQECAT Unveils First Probabilistic Catastrophe Model for U.S. Winter Storms

EQECAT Inc., an authority on extreme risk modeling announced the release of the first Winter Storm Model to quantify and manage insurance exposures associated with snow and ice damage in the continental United States. John P. Woods Co., Inc. (JPW), a reinsurance intermediary, and subsidiary of Arthur J. Gallagher &
Co. collaborated on the model development.

The EQECAT Winter Storm Model is designed to enable portfolio analyses for residential, commercial and industrial risk types. It covers a wide range of winter storm events, including blizzards, ice storms, Eastern Rockies, Northeast (Nor’easter), and Pacific Coast winter storms, and a variety of damages, including roof damage, pipe freeze, and ice damage.

“Winter storms have been causing annual losses in the lower layers or below the catastrophe reinsurance cover retentions of most companies, causing billions of dollars in losses to the insurance industry,” said Lou Estrema, senior vice president, JPW. “While individually these losses don’t generally impair a company’s ability
to write business, they certainly erode their property loss ratios.”

Historical industry loss data, such as that provided by Property Claims Service (PCS), show that winter freeze, snow and ice damage has accounted for approximately 8.5 percent of the total accumulated U.S. catastrophe losses since 1949.

“We believe that industry winter storm losses may be understated by as much as a factor or two due to under-reporting. The reason for this is that typically many of the claims come in long after industry sources have tallied losses for the particular storm. This is among the reasons why we believe the industry needed a model to better manage the exposure, and to make more informed risk transfer choices,” said Estrema.

The EQECAT model is based on detailed historical storm data, starting from 1961 and continuing through 2003. All the data was collected from the National Weather Service through hourly weather observations and local climatological reports. The vulnerability functions for the model were developed using 20 years of claims data from several insurance companies, and PCS historical data.

The EQECAT Probabilistic Winter Storm model reveals that insured losses could exceed $5 billion should there be a recurrence of a variation of the March 1978 storm, which affected a significant part of the East Coast.

JPW selected EQECAT to develop a probabilistic model that would help insurers to quantify and manage their exposure to these losses because “we believe that they are the technology leader and the innovator in natural hazard modeling,” said Estrema. “In addition, we have confidence in EQECAT’s ability to deliver the capability we need.”

“Through the use of sophisticated stochastic simulation and stratified
sampling, we have captured the full complexity and variability of winter storms,” added Dr. Mahmoud Khater, EQECAT’s chief technology officer.

“Winter storms are complex phenomena, which depend on a large number of parameters, such as snowfall amount, ice accretion, wind speed, and pre-existing snow depth, to compute the potential damage and insured losses. The EQECAT Winter Storm model takes into account the full range of parameters and has more than 12,000 stochastic events, producing very credible results,” said Khater.