The word prediction brings up a lot of negative connotations. We often laugh at predictions made about who will win an election or a Super Bowl, but insurers aren’t laughing at the predictions they are able to make about customer behavior that allows them to draw up models that limit exposure to risks.
In a recent report from Towers Watson (“Predictive Modeling Proving Its Worth Among P&C Carriers”), Brian Stoll, director and co-author of the report for Towers Watson, explains the value insurers have gained.
Stoll reports 70 percent of insurance carrier respondents use predictive modeling for personal auto and Stoll feels that figure understates the market share of carriers using predictive modeling.
“We didn’t weigh by market share,” he says. “If you start with the top 10 carriers, they all model and have been for years. If you look at this from a market share standpoint you probably have 95 percent doing [predictive modeling] and the rest are planning on it.”
Those not doing any predictive modeling for personal lines automobile are likely mutual companies operating in one or two states, explains Stoll.
“It’s tough to get around not modeling in big states,” he says. “You might be able to get away with it in smaller states.”
The terms predictive modeling and predictive analytics sometimes get thrown around interchangeably and Stoll isn’t sure there is much of a difference.
“Predictive analytics is probably broader in that modeling implies some formal mathematical treatment, where predictive analytics can be a series of metrics and measures that help you make better decisions,” he says. “Whether you call it predictive analytics or predictive modeling, you are getting at the same thing. They both shoot for the same goal but approach it in different ways. Personally, I would say predictive modeling is a subset of predictive analytics, but I’m sure you’d get many more definitions. The terms are becoming interchangeable.”
Stoll believes predictive modeling got its start in the personal auto line with the use of credit checks, a variable that had never been used before.
“It was not directly relevant to personal auto, but it is a good proxy for responsibility and responsibility matters,” says Stoll. “[Carriers] brought in statisticians and began looking at interactions of variables that historically had been treated one at a time. They started looking at other external and internal variables. It all combined to make predictive modeling what it is pre-telematics.”
Telematics is the game changer for predictive modeling, though, according to Stoll. Carriers moved from some meaningful proxies to actual information about the driver, how they drive, when they drive, and how much torque is in the vehicle.
“That information replaces a lot of proxies that created controversies from a regulatory standpoint and some discomfort,” says Stoll. “It enables better pricing because you have [driving] information on the drivers.”
The private view of a driver makes it difficult for another carrier to take that customer away without assuming the original company has correct information and the competition is going to match the first price.
“The potential is huge,” says Stoll. “The logistics are not easy so it’s taking a while for companies to get their arms around it. The big national companies are already there in different ways, shapes, and forms. Progressive is well ahead of everybody else. It is definitely coming and coming fast.”
It’s easy for a carrier to use a GPS device in a private automobile to capture all the information and produce scores on the driver’s driving habits. For commercial fleet operators, the value is different. The devices not only manage driver behavior, but pass on data about driver logistics and fuel efficiency.
“Good drivers save 10 percent of fuel compared to drivers that brake quickly or accelerate quickly,” points out Stoll. “That can be a huge number for a fleet.”
The telematics applications for insurance carriers with personal auto are maturing slower than in the commercial lines because the needs of the personal lines insurers are broader and needs to be integrated more efficiently to make effective use of the data, according to Stoll.
For commercial lines, the rating plans are class based. Carriers are not always thrilled with their exposure bases or class plans, according to Stoll.
“[Carriers] recognize there are big gaps between the ability of those tools to effectively manage exposure at the individual risk level and translate it to effective pricing,” he says. “Historically, they have given underwriters more discretion for that individual risk. Stability of operations, and stability of ownership are relevant and a lot of internal data is relevant, too, starting with interaction of variables carriers already have in house. Throw in the next layer—loss control and premium audit because the diversity of claims is greater than with personal auto—and all that information is relevant.”
Stoll points out that with personal auto there are many vehicles and each exposure is small, so the traditional methods of actuarial credibility work well. For commercial lines it’s much tougher.
“If you move to the extremes where you are writing $10 million limits the game changes from expected loss to frequency avoidance,” says Stoll. “If you can avoid even one loss through predictive modeling you can save $10 million. That’s a pretty good investment to make.”
Carriers have to determine if they are going to make better decisions with this information or without it? Stoll claims he has yet to see an insurance line where there wasn’t some significant lift from modeling.