Cat Surprises Create Model Questions

A spike in severe thunderstorms — coupled with greater-than-expected losses from Hurricane Ike — made 2008 a costly year for U.S. property/casualty insurers.

While industry representatives argue that they are well-capitalized to absorb the losses, some are growing concerned that catastrophe models are becoming less predictive.

“We haven’t been caught by surprise the way others were caught by surprise,” said Ernst Rauch, head of Munich Re’s Corporate Climate Center.

Rauch and other catastrophe experts made their comments during a online webinar sponsored by Munich Re on Thursday.

Munich Re estimates that 2008 will go down as one of the top five loss years in U.S. history, with insured losses topping $30 billion. There was a record of 10 significant natural catastrophes in the United States last year that caused $56.5 billion in overall losses and $25.1 billion in insured losses.

Insured losses could eventually rise another $5 billion as insurers adjust for existing claims.

Weather perils were the main driver of insured losses in the U.S., with thunderstorms causing record damages. Insured losses from severe thunderstorms topped $10.6 billion out of overall losses of $14.6 billion, according to Munich Re.

There were also 125 fatalities in the U.S. due to severe thunderstorms.

Beyond spurring thunderstorm activity, many insurers were caught off guard by Hurricane Ike, which made U.S landfall at Galveston, Texas in early September. Hurricane Ike turned out to be the third most costly hurricane in U.S. history with loss estimates currently at $15 billion.

Some insurers point to Hurricane Ike as an example of how catastrophe models fell short in 2008.

“Initial indications for Ike were $7 to $12 billion and the subsequent indications are at $12 to $20 billion in just a matter of months. That is a very significant departure,” said Franklin Montross, chairman and CEO of General Re, during a separate conference earlier this week.

One explanation for catastrophe models’ weak showing during Hurricane Ike is the unusual nature of the storm, said Carl Hedde, head of risk accumulation for Munich Re America, during the webinar. “Ike had some unique characteristics and differences from many other storms, ” he said, pointing to its large windfield, unexpected storm surge and its subsequent combination with a cold front in the Midwest that caused massive flooding.

“Cat models are just one of the tools,” Hedde added. “They are are a little weaker [at] predicting individual events.”

Insurers and reinsurers need to combine external catastrophe models with their own staff analysis and local knowledge to have predictive analysis work correctly, said Rauch. “We have always combined output with the vendor models along with our own expertise,” he said.

However, with one of the most surprising loss years behind them, U.S. insurers may be rethinking their approach to catastrophe models.

“When one looks at the emergence of cat models over the past two decades and how reliant this industry had become, there is a tendency that we really need to understand,” Montross said.


Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Risk Market News.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.