By MATTHEW BRODSKY, senior editor/Web editor of Risk & Insurance®
It's tricky enough to properly underwrite your more "glamorous" catastrophes, like earthquakes and hurricanes. These are the big-ticket items whose insured losses can tally into the tens--or in worst cases--hundreds of billions of dollars. Then you have floods. Not only are these events more "mundane"--far more frequent and often less severe than other CATs--they are a devil to understand.
"Flood is probably one of the more difficult hazards to predict and underwrite," said Tim Rose, senior vice president and chief underwriting officer for national accounts, property division, at Liberty Mutual.
One issue is that the best tools to determine flood risk--the flood maps from the Army Corps of Engineers--are products of "not an exact science," as Ralph Tiede, vice president of property risk engineering at Liberty Mutual, described them. Flood areas are constantly changing, and any redrawing of the maps can be controversial. After all, having a property located in a designated high-risk zone can affect the affordability and availability of insurance.
Commercial insurance underwriters work with flood maps in analyzing a property's exposure as high, moderate or low, according to Rose. These designations help underwriters to set a flood sublimit in a commercial property policy. A lot of today's commercial flood insurance coverage is sublimited, Rose said.
Still, what flood observers will tell you is that flood exposure isn't as simple as looking at the Army Corps' maps. One 100-year flood zone isn't the same as another 100-year zone, explained Abebe Jemberie, the senior research engineer at AIR Worldwide, who is responsible for developing the catastrophe modeling firm's new U.S. flood model.
"Flood is very, very location specific," he said.
So much so that one property in the same 100-year flood zone could have a totally different flood exposure than another property, even if the two are a mere 10 or 20 yards apart. During a flood, one property could be submerged and the other dry, Jemberie explained.
WHAT IS TO BE DONE?
Corporate risk managers at large organizations with big budgets can pay to have a professional engineer visit flood-exposed sites and provide a detailed assessment of the risk, Rose said.
The risk manager and his or her broker can then offer that data to underwriters, who would then, the risk manager hopes, better price the risk.
Without such on-the-ground information, underwriters can also rely on flood models like the one AIR is working on for the United States. AIR already has flood models for European countries, as does fellow modeler Eqecat Inc. Competitor Risk Management Solutions Inc. also has developed flood models for some European countries, like Germany and the United Kingdom.
Jemberie indicated that clients--meaning mainly insurers and reinsurers--have demanded a U.S. flood model.
Flood models--of course, like all catastrophe models out there--won't be perfect for forecasting losses for given probabilistic events, but they can at least give numbers that underwriters can work with. The model could take into account the design of levees and the chance of their failure based on that, according to Jemberie. They could incorporate information on the potential flood damage inflicted by a failed dam based on how much water the dam held back at time of failure and on how much property value was downstream. And for well-known and better understood levee systems that are at high risk--such as those on the Mississippi and Missouri rivers and around Sacramento, Calif.--the model could use specialized data already collected by experts.
May 1, 2011
Copyright 2011© LRP Publications