Car sales are thriving, and so is the likelihood of fraud in loan applications, says a San Diego company specializing in fraud detection.
PointPredictive, with its Auto Fraud Manager software, has developed what it calls pattern recognition models that can both identify risk and potentially cut fraud and early payment default losses in half.
“We estimate the level of annual fraud risk for 2017 to be between $4 billion and $6 billion,” says Tim Grace, CEO of PointPredictive, which launched the software in January.
PointPredictive provides a customer, typically a lender, with a three-digit score for three components of the loan application: misrepresentation, dealer risk and early payment default risk. The numbers run from one -- the lowest risk -- to 999.
The scoring is available to the lender in a matter of seconds. “Lenders can choose their threshold of risk,” says Frank McKenna, PointPredictive’s chief fraud strategist. Each score includes up to three reason codes to explain the rating and how the risk might be treated.
“You can’t trust a credit score,” McKenna says. “It is not a good indication of risk.”
While lenders are the obvious target market for the analysis, dealers, too, would benefit from the perspective it can offer, McKenna says.
“We think eventually dealers will want to see all of the risks on an application so they can gauge if there is any significant fraud risks on the loan before they send it to a lender to review,” he says.
While loan application formats may differ, PointPredictive says it can read data from any system. Specialists map data from various formats into a common format, McKenna says.
Terry Dortch, CEO at Automotive Compliance Consultants in Crystal Lake, Ill., says credits scores are the “best way to tell whether or not a person pays bills on time.” Still, he says, “There is not a lot of predictive software out there,” and PointPredictive’s approach appears new.
While most loan operations would have their own system for reviewing applications, not many are likely to have anything like Auto Fraud Manager. “It would present itself well to someone that did not want to reinvent the wheel,” he says.
He adds that it’s not that unusual for dealership staff to alter credit applications.
“It is more common than you would think,” Dortch says. There are still dealers out there who employ “less than scrupulous” staff eager to collect the fees.
How many bad apples are there in the bushel? PointPredictive says that within some lending portfolios, as few as 3 percent of the dealerships were responsible for 100 percent of the fraud and early payment default risk for those lenders.