Survey: P&C Insurers Expand Usage of Predictive Modeling

February 6, 2012

Predictive modeling continues to gain momentum among North American property & casualty (P&C) insurers, according to findings from global professional services company Towers Watson’s third predictive modeling survey.

The findings show that while the potential to advance critical areas of P&C pricing and risk assessment through predictive modeling is not as mature in commercial lines as it is in personal lines, commercial lines adoption will continue to accelerate over the next two years.

The survey revealed that nearly all (97 percent) U.S. personal lines respondents said they view sophisticated underwriting and risk selection as essential or very important. As such, the majority of U.S. personal lines insurers are already committed to predictive modeling, with approximately 85 percent saying they use or are planning to use it.

While standard commercial lines insurers have lagged their personal lines counterparts in recognizing the importance of moving to adopt predictive modeling, those carriers are now actively adopting the technique, with roughly 70 percent indicating they either currently use or plan to use it in underwriting, risk selection, rating and/or pricing within the next two years.

“The optimism expressed by the senior executives responding to the survey suggests the range of future uses for predictive modeling is broad, and will include not only pricing and product innovation but also new refinements in areas such as underwriting, risk selection, claim applications and target marketing,” said Brian Stoll, Towers Watson director and the survey’s coauthor.

“There is a good reason for optimism: An increasing number of survey participants that use predictive modeling report it is improving both their top- and bottom-line results. This profitability improvement encourages P&C insurers to find even more ways to extend predictive modeling applications and further access the benefits it can offer.”

While standard commercial insurers currently using predictive modeling are in the more modest 25 to 41 percent range (indicating significant growth from a year ago), those planning to start programs are in a more robust 31 to 48 percent bracket.

Part of the motivation for the aggressive plans of commercial insurers is likely explained by concern they expressed in survey responses about the ability of the standard industry class plans and exposure bases to appropriately segment individual risks, in that predictive modeling provides the path to more accurate pricing.

The greatest growth potential appears to be concentrated in the commercial property, commercial multiperil (CMP) and business owner policy (BOP) lines of business, where nearly half (48 percent) of respondents indicate they intend to implement predictive modeling for risk selection and/or pricing. Commercial automobile carriers are also showing interest, with 37 percent responding that they plan to use predictive modeling.

Survey findings also show that the use of predictive modeling is, at least in part, driven by the importance that carriers place on sophisticated underwriting, risk selection, rating and/or pricing.

Insurers using predictive modeling also said it is pivotal in achieving both strong top- and bottom-line results. Bottom-line benefits of rate accuracy, loss ratio improvement and improved profitability all received positive responses of nearly 75 percent or more from survey participants. Furthermore, over a third to nearly half of respondents also cited additional positive impacts on the top line from expansion of underwriting appetite, improved renewal retention and increased market share.

“What’s important to note is that these percentages are up about 10% to 20% over last year’s results, indicating the enduring sustainable benefits of predictive modeling,” said Klayton Southwood, a senior consultant at Towers Watson.

Leading personal auto lines insurers that committed to predictive modeling early in its development are actively adding depth to their programs by introducing or expanding their use of telematics, the technology of sending, receiving and storing information via telecommunication devices. Currently, personal lines respondent carriers that use telematics are focusing on only a few areas, including measurement of annual mileage, tracking how and when a vehicle is being driven, who is driving the vehicle and where it is being driven.

More expansive plans are under way, according to the survey, with 89 percent of personal lines respondents that either currently use or plan to use telematics planning to use the data prospectively in rating, and 83 percent planning to use it to provide information to insureds to help improve driving behavior.

Commercial carriers’ use of telematics is focused on the same four areas mentioned earlier: 87 percent of commercial lines respondents that currently use or plan to use telematics have plans to track mileage, where and when the vehicle is being driven, as well as who is driving the vehicle, the survey indicated.

Sixty-nine executives from U.S. and Canadian P&C carriers participated in Towers Watson’s third web-based survey on predictive modeling, including chief actuaries, pricing officers, claim officers and other senior executives.

The survey focused on how predictive modeling is being used to support rating/pricing and underwriting/risk selection plans at their companies, and was conducted in October and November 2011. Responding companies represent a significant market share of U.S. personal lines carriers and commercial lines carriers. Thirty-five percent of U.S. respondents had 2010 annual direct written premium greater than $750 million.

Source: Towers Watson