Updating Your Models for Hurricane Season

The North Atlantic hurricane season has begun, and CAT models need to be updated to remove the possibility of major losses.

June 1 opened the North Atlantic hurricane season, with this year marking the 10th anniversary of one of the costliest storms to make landfall in the U.S. — Hurricane Katrina. Each year, hurricane season puts catastrophe (CAT) models to the test, with potentially millions of dollars riding on their accuracy. The loss estimates calculated by CAT models can play an important role in protecting your organization from financial loss. The models have changed a lot over the past several years. For example, Hurricane Andrew in 1992 exposed the shortcomings of traditional actuarial methods that insurers use to model risks. And the billions of dollars in insured losses from Hurricane Katrina in 2005 helped lead to today’s CAT modeling rigor and its universal acceptance and use by the industry. New Storms Change CAT Models CAT models use algorithms to estimate potential losses stemming from a catastrophic event. Over the 10 years since Katrina, CAT modeling has become more complex because of technology improvements and the greater availability of data. After a significant storm, the models are updated based on the new data and a larger body of knowledge. These changes could considerably affect your property insurance and risk management strategies. Here are some CAT modeling factors — which for U.S. hurricane exposures have changed several times in the last few years. You should consider the items below as you prepare for this year’s hurricane season:
  • Check your policy, including deductibles, coverage limits and sublimits, to ensure they’re adequate and realistic; check that exclusions are acceptable.
  • Ensure the quality of your CAT modeling data. Incomplete data causes more uncertainty for insurers; improving the data enables more accurate loss estimates and reduces the uncertainty for the underwriters.
  • Take a big picture view of your CAT exposures. By modeling your worldwide portfolio, you can identify regional drivers, which can help put U.S. hurricane risks in perspective. Also, using actuarial resources after a CAT or non-CAT claim can help evaluate your organization's total cost of risk (TCOR), which can better inform how you use your risk management resources.
If you have locations in CAT-prone areas, you can fine-tune their CAT loss estimates with an understanding of how they’ve changed with each model update. Aligning your risk data with CAT modeling changes can yield better outputs for insurers to underwrite your risks. To register for a webinar on June 17, 2015, on the lessons from Hurricane Katrina, click here.

Cheryl Fanelli

Profile picture for user CherylFanelli

Cheryl Fanelli

Cheryl Fanelli delivers best-in-class analytics for modeling catastrophe exposures for clients. As manager of Marsh's CAT Modeling Center of Excellence, Fanelli and her team work closely with consultants and brokers to review, analyze and model client property data against models of historical or potential catastrophic events.

Read More