How auto lenders can improve credit opportunities

Major automotive lenders are relying on "alternative data" beyond a traditional FICO score to make or refine credit decisions. However, there's still plenty of room to grow, according to LexisNexis Risk Solutions.

Further adoption could improve credit opportunities for numerous dealership customers, according to the data provider.

Solomon Semere, senior director of consumer credit and marketing strategy for LexisNexis Risk Solutions, estimated 50 million consumers have thin or nonexistent credit files.

Through alternative data, "you can get visibility into nearly all of them," he said during a Dec. 9 webinar hosted by artificial intelligence underwriting provider Scienaptic AI. "Upwards of 90 percent of those consumers now become scoreable."

Reevaluating such customers using alternative data can bump auto loan approval rates by 15 to 20 percentage points, according to Kevin King, LexisNexis Risk Solutions vice president of credit and marketing strategy.

Alternative data also allows a lender to refine terms on the borrowers it would have approved using a traditional score, King told Automotive News. For example, recognizing a FICO 740 score really carries the risk of a FICO 780 and deserves a deal adjusted accordingly. Conversely, other consumers are revealed to represent a greater risk than traditional scoring indicates, according to Semere.

Auto lenders have been more bullish on alternative data than other lending segments, according to King. But there's still a "huge" opportunity for expansion in the primary-use cases of underwriting and prequalification, he said. King estimated no more than half of the Top 25 captive and noncaptive lenders used alternative data for underwriting or prequalifications.

About 40 percent of LexisNexis' auto lenders who use alternative data apply it to every applicant, King said. Another 40 percent of the adopters use it to reevaluate applicants who fail to qualify for a loan or who have insufficient credit history, he said. The remainder use it to refine pricing on applicants who already qualified, he said.

"It probably would behoove a dealership to really understand what your lender partners are doing [with alternative data] to be able to score a thin file or a new-to-credit consumer, so that the dealer can help arrange the credit for a consumer that could be creditworthy," Satyan Merchant, TransUnion senior vice president of auto financial services, told Automotive News.

King said adoption of alternative data appears proportional to the competitive pressure of the lender's segment. More direct lenders use it than indirect ones, and more noncaptive lenders consider it than captive finance companies.

"That's changing," King said.

The largest captive providers, which King estimated were typically the least aggressive on alternative data, are beginning to intensify their usage. Nearly every digital auto lender uses alternative data, he said.

King defined alternative data as any information not consistently present within a FICO score. The use of this information must pass regulatory muster as well as common-sense scrutiny, and be disputable by the customer in the case of an error, he said. The details also should be something a consumer would expect to be considered.

For example, it would be inappropriate to use a borrower's social media posts to evaluate that individual's credit risk, he said.

LexisNexis evaluates that information and issues consumers a RiskView score it calibrates to FICO, King said. For instance, a 650 RiskView should predict the same degree of creditworthiness as a 650 FICO, he said. But the company is trying to complement FICO, not replace it, he said.

King said the automotive loan approval rate using traditional means hovered in the mid-40 percent range between January 2020 to October 2021. Applying a RiskView alternative data score to the populations denied for insufficient credit history or a low traditional credit score would yield an approval rate in the low 60 percent range, according to King.

He estimated about 25 percent of this improvement involved the credit-invisible population, and the remainder involved those whose creditworthiness had been underestimated.

Alternative data include factors such as liens, business ownership and attendance at a two- or four-year college, according to Semere.

Equifax on Dec. 20 reported success with another type of alternative data that it planned to bring mainstream. The credit bureau plans to incorporate "Buy Now, Pay Later" results into credit reports for consideration, calling itself the first agency to standardize this process. The term refers to buying goods on interest-free installment plans.

The BNPL data gave customers with "thin" (two or fewer entries) or "young" (two years or less of credit history) credit files a 21-point gain in FICO. It also helped consumers with previous late payments improve their scores 13 points.

King called two categories of alternative data particularly relevant for auto loans. The first involved credit-seeking behavior not visible to credit bureaus, such as applications for online and subprime loans and checking accounts.

The other involved what he called "credit life event data," something he said was "really impactful in the auto space." This included professional licensing and stable rental history, King said.

Some consumers experience the traditional credit-building history that starts with a first bank account during childhood, CU 2.0 founder Kirk Drake said during the Scienaptic webinar.

"That traditional path doesn't necessarily exist for a huge portion of our population," said Drake, who earlier observed that a percentage of consumers don't see a difference between a bank account and a prepaid debit card.