Predictive Modeling Leveling Commercial Insurance Prices

Effective predictive modeling enhances underwriting and pricing profitability. This is Towers Watson’s conclusion after analyzing results of its “Commercial Lines Insurance Pricing Survey” (CLIPS), which compared prices charged on policies underwritten during Q2 2010 to the prices charged for the same coverage during the same period of 2009.

According to survey results, commercial insurance prices in aggregate declined by 1% during Q2 2010, and insurers that utilize predictive modeling in their pricing and risk selection process reported that they were better able to hold price levels. While data for most lines indicate flat prices, the commercial property, directors and officers liability, and employment practices liability lines show price reductions.

“The commercial lines insurers that are taking advantage of predictive modeling are finding new rating variables and sources of data, and are applying the results in new and innovative ways,” says Bruce Fell, practice leader of Towers Watson’s risk consulting and reinsurance brokerage services to the P&C industry in the Americas.

Yet those companies that use predictive modeling techniques for pricing and risk-tiering reported slight price increases—on average—compared to companies that do not use predictive modeling. Insurers that have yet to implement predictive modeling experienced price decreases greater than the overall decline of 1.0%. The most recent survey was the first time Towers Watson queried respondents—representing a cross section of U.S. P&C insurers—on their use of predictive modeling and, while these indications are preliminary, they point to the power of predictive modeling.

CLIPS results indicate that accident-year-to-date 2010 loss ratios deteriorated 4% relative to year-to-date 2009. This deterioration, which is based on six months of information and is, therefore, preliminary, is consistent with an estimated deterioration of 4% for accident-year 2009 over 2008. The firmer prices in year-to-date 2010 on an earned basis are offset by somewhat higher claim cost inflation indications than that observed in 2009.

For reprint and licensing requests for this article, click here.
Core systems Data security Data and information management Policy adminstration Security risk Analytics
MORE FROM DIGITAL INSURANCE