Auto lenders fear worsening economy could lead to more fraud

Auto lenders polled by artificial intelligence and data provider Point Predictive for its 2023 Auto Lending Fraud Survey fear worsening economic conditions could push loan fraud and defaults higher this year.

The survey showed 70 percent of auto lenders are preparing for a declining economy this year compared with 2022. Another big concern was that fraud was on the upswing.

One in four lenders surveyed said they are not tracking fraud as soon as it occurred, with 35 percent of lenders saying their frontline employees — underwriters looking at the loans before they are funded — are not trained or prepared to investigate or identify fraud. Only about 30 percent of lenders reported using the U.S. government's fee-based electronic system called eCBSV to confirm borrowers' social security numbers are legitimate. The eCBSV launched in June 2020 to help stop synthetic identity fraud.

"There's a lot of credit-tightening and a lot more scrutiny happening," Frank McKenna, co-founder and chief fraud strategist for the company, said.

Point Predictive's data consortium has information on more than 134 million loan applications totaling $2.4 trillion from U.S. dealerships and lenders.

"Lenders are looking at the highest quality loans that they can book, so they're being a lot more careful," McKenna said. "That's why you see less volume at lenders, maybe a flight to quality as the market turns. Fraud is part of that."

Income misrepresentation was the biggest fraud concern, the survey showed. Synthetic identity risk and dealer fraud also were worries. Pay stub forgery is still a problem for auto lenders, the majority of whom said they believe as much as 10 percent of pay stubs are false or fabricated.

Early payment default on auto loans indicates origination fraud, according to 91 percent of survey respondents. An early payment default occurs when a loan defaults within six months after the borrower buys and finances the car, McKenna said.

More than half of lenders surveyed this year said dealer-perpetrated fraud is a serious concern. Last year, 10 percent of lenders said they stopped working with 50 or more dealers because of fraud, Point Predictive said.

The four main types of dealer fraud, according to McKenna, are high levels of identity theft; income fraud; power-booking, a term describing a dealer inflating a car's value by listing options that aren't there; and high levels of loan default.

McKenna said three things dealers can do to help prevent fraud are train frontline staff, align incentives to ensure salespeople and finance managers are not just incentivized to make sales but to stop fraud, and use technology.

The findings are from Point Predictive's December 2022 and January 2023 survey of 38 risk management executives at auto lenders, banks and finance companies. Those responding to the survey included more than 35 lenders representing subprime to prime originations and captive and indirect providers.

Lenders are responding to this year's fraud risk concerns by increasing internal analytics and purchasing new technology.

Darren Schlosser, sergeant for the Houston Police Department auto theft division's vehicle fraud unit, told Automotive News that his unit has arrested 147 fraudulent car buyers since 2017 who were actively committing fraud inside dealerships. Those arrests prevented more than $7.3 million in fraud.

Schlosser works with Houston-area dealerships to arrest fraud perpetrators and travels across the U.S. training other law enforcement and dealerships on how to identify consumers attempting document fraud, identity theft and synthetic identity.

"Last year, we did 125 investigations that yielded $8.8 million," Schlosser said of total fraud investigations in the Houston area.

Most of Schlosser's investigations are at new-car dealerships, he said.

"Dealers are so interested in fraud right now," McKenna said. "I've never seen anything like it. I think because dealers are getting hit so hard by these identity thieves most dealers are looking at doing something, training or technology or something."