Predictive modeling is rapidly changing the market landscape for the small commercial package risks, and even making a difference for underwriters of large professional liability accounts, carrier leaders said at recent insurance conference.
Executive SummaryPredictive modeling is rapidly changing the market landscape for the small commercial package risks, and even making a difference for underwriters of large professional liability accounts, carrier leaders said at recent insurance conference.
“There are winners and losers being created today in this space,” said Mary Hennessy, a former senior executive for GMAC Insurance, Overseas Partners Ltd., and TIG Holdings, who has also served on the board of directors of several insurance carriers. The winners, she said, are successfully using predictive analytics in their commercial lines businesses.
Her comments came during a session of the St. Joseph’s University Insurance Financial Symposium last year, where property/casualty carrier executives and board members also discussed the appearance of underwriters—and underwriting committees—on boards of directors. In addition, session moderator Eric Simpson, a senior vice president for Towers Watson, asked the panelists to offer views on an underwriting-related development outside the boardroom—the integration of predictive analytics into underwriting and pricing decisions.
Mary Hennessy, insurance company board member and vice chair of EagleEye Analytics
Hennessy, who now serves as vice chair of the board of EagleEye Analytics, a provider of predictive technology for the p/c industry, said that predictive analysis is rapidly changing the market landscape for the smaller commercial risks.
Drawing some parallels to historical developments on the personal lines, she said: “When I started in the business, there were personal lines underwriters that made accept and reject decisions on each or every risk, or decided what company [within the group] am I going to slot the risk in and what’s the rate going to be.”
John Baily, insurance company board member and former PwC audit partner
Today, there are no personal lines underwriters other than in the high net-worth business. And there is a whole lot of business in the $10,000-and-under commercial category that is going to go exactly in that direction.”
“It has well started, and I firmly believe that the [commercial lines] evolution is not going to take 20 years like it did when you look at the catastrophe space, or when you look at where Progressive and GEICO are today in personal lines.”
“It’s going to happen in the next three or four years,” she predicted.
Philip Bancroft, chief financial officer of ACE Group, said there is also “some thinking emerging” that suggests predictive modeling will be helpful in some large-account commercial lines business, such as directors and officers liability and professional liability.
“Not the be all and end all, and not something that you just would plug into a machine and get the answer,” he said. But factors, such as the length of time management has been in place, the credit quality of the organization, the quality of the board could all be important considerations, he suggested. “Maybe they’re not traditional predictive modeling measures like a credit scores, [but] certainly, we and others are studying how to apply predictive modeling in this much broader and much bigger commercial lines space,” Bancroft said.
John Baily, who serves on the boards of several property/casualty insurers, is not keen on the idea of D&O modeling. “I get a little scared,” he said. “Sometimes, we’ve got to use some common sense too,” said Baily, who is a former audit partner for PwC. “As an audit committee chairman for the last 12 years, I have never ever been talked to by an insurance company writing D&O,” he said. “So I don’t know how they develop [an assessment of] what the quality of the board is or whether the board really has oversight or not.”
“I would be very interested in what model they’re using,” Baily said.
Hennessy agreed with Bancroft that predictive modeling is taking hold for all risk sizes, and that analytics could be used in professional liability and D&O to narrow down underwriting criteria. “Maybe the underwriters are thinking there are 60 significant things, but there are really only 20 because the other 40 are proxies for those 20. A good predictive analytics tool can find those, find what are the right questions…, and develop scores that underwriters can use to be more effective and efficient in what they’re doing.”
“I think it’s going to revolutionize the industry, not just on the pricing side,” she said, going on to give the example of commercial property risks, where it will no longer be necessary to pay for inspections on many properties. “And on the personal lines side, we don’t need to order MVRs [motor vehicle reports] for certain risks. The models are showing this consistently over tens of thousands, if not hundreds of thousands, of exposures,” she said, highlighting the efficiencies that can be gained to lower expense ratios.
“I am upbeat about what’s going to happen in this space as long as companies don’t fall behind. Those companies that do are going to find themselves pretty well out of the game in a decade,” she said. “We will never ever replace good underwriters, but we can make the whole process more efficient,” she said.