More than three-quarters (76%) of personal lines carriers surveyed recently said they view sophisticated predictive modeling techniques and tools as essential to pricing products more accurately, according to a survey conducted by
However, while only 56% of standard commercial carriers said it was important, that group of insurers is taking a closer look at predictive modeling’s advanced statistical modeling techniques that use critical data related to individual policyholders, competitors, marketplace conditions and customer behavior, to determine optimum pricing levels.
“As personal lines have paved the way, we believe commercial lines will move faster and with more effectiveness, given the lessons learned and the current environment,” said Klayton Southwood, survey co-author and Towers Perrin senior consultant. “The predictive modeling trends we’ve seen in property/casualty are consistent with those beyond insurance, where predictive modeling is growing in such diverse areas as government oversight, health care, education and myriad others.”
Ninety insurance executives from 81 property/casualty carriers participated in the Web-based survey, which focused on perspectives on how predictive modeling is being used to support decisions related to rating, pricing, underwriting and risk. The survey was conducted from April 21, 2009 through June 6, 2009.
The study shows that current use of predictive modeling varies significantly by business segment—led by personal auto at 68% and homeowners at 42%—but many commercial lines carriers said they are investing in predictive modeling across all segments, including commercial auto (44%), workers compensation (43%) and commercial property (36%).
These findings indicate a dramatic evolution in insurers’ pricing strategies as, historically, many firms relied solely on traditional cost-based pricing approaches to set prices. While cost is still a critical consideration, robust modeling techniques are helping carriers reflect an ever-changing competitive landscape and customer behavior in their pricing structures.
When asked what primarily drove their firm’s use of predictive modeling, 51% of respondents indicated profitability considerations. The secondary reason cited most often was competitive pressure (22%). Other reasons include growth considerations, good business practices and regulatory changes.
Further, a majority of carriers said they are using, or plan to extend their use of, predictive modeling beyond rating/pricing and underwriting/risk selection, most notably in the areas of claims (65%), catastrophe exposure (62%) and marketing (51%).
“The vast majority of personal lines carriers see rating and underwriting sophistication as critical -- or at least very important -- as our research shows that there exists a strong relationship among rating, pricing, underwriting, risk selection sophistication and financial results,” said Brian Stoll, Towers Perrin senior consultant and co-developer of the survey. “The synergies that exist between personal and commercial lines suggest that predictive modeling will quickly become more pervasive among all types of carriers.”
Among other key findings from the survey:
• Challenges associated with data (29%) lead the list of hurdles for increasing rating or underwriting sophistication, followed by production systems (20%) and cultural issues (15%).
• When it comes to predictive modeling sophistication (an aggregate of several factors, including competitive sensing, modeling approach, and underwriting and risk selection, among others), personal auto carriers lead the field, while homeowner carriers have become increasingly sophisticated. On the commercial side, commercial auto carriers have increased their collective sophistication around pricing, but still rely heavily on traditional rating plans and factors.
• Most medium and larger carriers have full-time predictive modeling units devoted to rating/underwriting modeling support. Small carriers tend to devote five or fewer full-time employees, while larger carriers dedicate at least six.
