Dive Brief:

• Upstart is launching an auto lending platform that will use much of the same alternative data model that helped it target borrowers the company found creditworthy even though they may not be considered prime based on their credit scores or credit history, the online lender said Wednesday.
• Upstart's auto model calculates the probability a borrower will default or prepay alongside the vehicle's residual value to generate a loan offer unique to each applicant, the company said.
• The platform syncs with DMV records, eliminating the need for consumers to find and enter their vehicle identification number (VIN) or license plate number. Banks also no longer need to manage paperwork regarding title transfer, lien placement or, with refinances, payoff of the borrower's existing loan.

Dive Insight:

Between 2017 and 2019, alternative data helped Upstart approve 27% more personal loans than traditional lenders with an average annual percentage rate (APR) that is 16% lower, the online lender told the Consumer Financial Protection Bureau (CFPB). Upstart received a no-action letter from the regulator in 2017, allowing it to use alternative data under monitoring from the bureau.

The lender came under fire in February after the nonprofit Student Borrower Protection Center published a case study indicating its model would charge a hypothetical graduate of Howard University, a historically Black university, about $3,500 more over a five-year repayment term than a similarly situated New York University graduate. The APR on the Howard loan is nearly 5 percentage points higher — 21.29%, compared with 16.34% on the NYU loan, the case study found. And the Howard borrower's $1,960 origination fee was $729 higher than that of the NYU borrower. Five Democratic senators sent a letter asking CEO Dave Girouard to prove the online lender tests its credit model for disparate impact, or disproportionate negative effects against a protected class of people.

Upstart's auto loan underwriting model looks at more than 1,500 factors to assess risk, Val Gui, the company's general manager, told American Banker. Borrowers can fill out an application in minutes via a mobile-friendly, bank-branded app. And the platform will present alternative loan offers from which a potential car buyer can choose.

Banks, too, have found hurdles in the auto lending space. JPMorgan Chase filed a patent to use blockchain technology to track the vehicles it finances. The bank said it can help prevent fraud by anchoring VINs to a blockchain.

Other banks, such as Wells Fargo, have backed away from some segments of the auto-lending market to cut down on risk. The San Francisco-based lender this month said it would no longer accept auto loan applications from most independent car dealerships. Independent car dealers, which typically sell used cars, account for less than 10% of the 11,000 businesses through which Wells Fargo sells auto loans. Franchise dealerships, on the other hand, focus on new vehicles from specific manufacturers.

Subprime auto-loan borrowers have also been the target of shady lending practices. Santander Consumer USA last month reached a $550 million agreement to settle charges from 34 attorneys general that it made auto loans it knew low-income borrowers could not pay.

"Auto came very naturally to us," Gui told American Banker. "It’s a big market, and there are a lot of people paying way too much for car loans."

Upstart did not divulge how much it will charge banks to use the auto lending platform, but the company’s co-founder, Paul Gu, said pricing is based on the number of loans originated.

"If we aren't able to help the bank actually convert many loans, then we'll share in the downside," he told American Banker.