Ask the question, "What's new with catastrophe modeling?" and you usually get back replies like:
"An updated storm surge component."
"A new windstorm model for Eastern Europe."
"Results tailored for specific locations or property schedules."
But lately you might notice a new reply. Besides listing their new products and geographic reach, risk modeling vendors are also talking about what's new in how their products are being--or should be--used. They say their products ought to be more fully integrated with carriers' underwriting systems.
It is one of the "next frontiers," said Paul VanderMarck, executive vice president, products, for modeling vendor Risk Management Solutions Inc.
CAT modeling software allows users to apply statistical data and the latest scientific understanding on hurricanes, earthquakes and other disasters to estimate their probability of losses should such an event befall their book of business.
Currently, said VanderMarck, only some carriers have brought modeling into their underwriting process, and those are mostly the large commercial specialty insurers.
But another executive said those that are using it are getting after it in a hurry.
"There are firms moving down this path quite rapidly," agreed Kenneth A. Travers, senior vice president at ABS Consulting/Eqecat Inc.
Travers and VanderMarck spoke at the catastrophe modeling seminar put on by the Reinsurance Association of America earlier this year in Tampa, Fla.
The aim, said VanderMarck, is for modeling to become enmeshed in the underwriting of more automated commercial lines, such as small business and middle market.
Carriers writing in these markets, and any other insurer without the modeling-underwriting integration, might only be using the catastrophe software after the underwriting fact as a double-check on their pricing and accumulations. Or they might rely on their brokers to run the models for them. Some carriers might not even use models at all; instead, says VanderMarck, they use modeling "proxies," such as distance to the coast or soil type, to check their underwriting.
One cannot blame the insurance companies for the situation entirely.
Said Travers, obstacles stand in the way of the achieving the next frontier, including the lack of a standard data format for modeling software, the difficulties in accessing needed data, and the simple fact that the technologies used in underwriting workflows can't always talk to the modeling software. Modeling software often has different data constraints than underwriting systems.
Still, in the next five years, VanderMarck foresees much more dynamic integration of models into underwriting, and a greater granualization (aka, a more accurate picture) of exposure because of it.
Underwriters can also expect improved efficiency, a much more customized modeling experience and increased access to market data, according to Travers.
"This is an area that the market has to go," said Travers.
MATTHEW BRODSKY is senior editor/Web editor of Risk & Insurance®.
April 15, 2008
Copyright 2008© LRP Publications