Thirteen lenders, representing more than a third of U.S. auto loan originations, shared internal auto lending trends and experiences to combat fraud at a meeting of PointPredictive’s Automotive Lending Fraud Consortium last month.
Lenders in the consortium agreed to meet quarterly to share the patterns of fraud in their data. Most of the lenders, including captives and noncaptives, are large, originating about $30 billion in loans and leases a year, said Frank McKenna, chief strategist and co-founder of PointPredictive.
During last month’s meeting in Dallas, the consortium’s first, the lenders defined fraud in auto lending as an intentional misrepresentation on loan applications, McKenna said. They discussed different types of fraud and the kinds of patterns they should routinely share with one another to establish a strategy to address fraud and build best practices as an industry.
PointPredictive has pledged to support the consortium by collecting lender data, building a predictive application for risk scoring and dealership scoring, providing Ph.D. scientists and fraud experts to analyze data and organizing the meetings.
The company estimates the level of annual fraud risk for 2017 is between $4 billion and $6 billion. “This is something that lenders want to address now,” McKenna said. “The fact that they’re getting together like this tells me [now is] the time.”
Ninety percent of the lenders said fraud was a growing concern. A year and a half ago, that percentage would have been lower, McKenna believes. With auto loan volume at a high and some lenders tightening their lending criteria, lenders are seeing how fraud can impact their performance, he said.
PointPredictive aims to work with 60 to 80 percent of the auto finance industry. “The more [lenders] you get, the more you can reduce losses,” McKenna said.
Since the announcement of the consortium last month, four more interested lenders have contacted PointPredictive, he said. McKenna previously worked on data and lending consortiums for mortgage and debit cards with BasePoint Analytics.
PointPredictive and lenders in the consortium typically identify patterns of income fraud when borrowers misrepresent their income.
“If you see a borrower reporting much higher income than the average in the consortium, [or] if you see income escalating over time unreasonably,” fraud may have been a factor, McKenna said.
The lenders discuss an individual dealership’s potential to be high-risk “when it looks like the dealership has submitted a high number of loans that have fraud on them in a short period of time,” McKenna said.
A new finance manager or fraud ring, for example, can cause a fraudulent loan increase, he said.
Fraud rings are becoming increasingly common, jumping from dealership to dealership, McKenna said. Typically, a few members of a ring will go into a store and look for a hole “like a finance manager who’s lenient.” If they find one, they will send someone every night to get as many cars as they can until the dealership or a lender becomes suspicious. After the dealership or lender catches on, the fraud ring will move on to another dealership, he said.
Especially with luxury vehicles, “rather than try to resell them here in the U.S., [fraud rings] will actually ship them to Asia,” McKenna said. “If they can get a Mercedes or Range Rover, they can go to Asia and sell it for a quarter of a million dollars or more.”