Point Predictive, CreditMiner tech helps dealers combat fraud
New technology from Point Predictive and a CreditMiner-TransUnion partnership allow dealerships to take a more active role in preventing borrower fraud.
BorrowerCheck for Dealers is “the first in a planned series of products for dealers” leveraged from the fraud analytics firm’s vast database of records including income reports, loan applications, fake employers and fraudulent loans. It is a resource previously only available to lenders.
Dealerships can check the customer’s name, address and Social Security number and see if more proof is necessary, according to Point Predictive. The company also has created a new phone verification mechanism.
CreditMiner’s new “Identify” software leverages TransUnion’s ability to check a customer’s driver’s license image against a selfie and the records on the person ostensibly described in the license. It also evaluates the credibility of the customer’s phone. Ken Luna, CreditMiner strategic partnerships vice president, said lenders had been using these TransUnion capabilities, but they weren’t prevalent among dealerships.
Earlier this year, Point Predictive said its fraud team found more than 16,600 suspicious auto loan applications in 2021, a 260 percent increase from 2020. These questionable 2021 loan applications sought a combined $309 million worth of financing for vehicles. And that’s just part of what Point Predictive called a potentially $7.7 billion issue facing the industry last year.
Frank McKenna, Point Predictive chief fraud strategist, on Sept. 6 told Automotive News nearly every dealership his company has talked to is reporting more fraud.
“They’re telling us that their fraud rates have increased pretty significantly,” he said.
Retailers who have never seen fraud before are reporting multiple cases — and those incidents primarily end with indirect lenders forcing dealerships to repurchase the loans, he said. Buying back a loan can set the dealership back tens of thousands of dollars, a loss that would require multiple car sales to cover, he said.
“Those dealerships now are getting hit with a lot more of those,” McKenna said. “They’re looking for ways to stop it.”
McKenna said these chargebacks arise when lenders discover identity theft, which he said has become more sophisticated. The fraudsters have “really good fake IDs” and “know everything about the customer,” he said.
This knowledge has reached the point scammers can answer verification questions such as their victim’s residence in a given year and last auto lender, McKenna said. He said fraudsters will obtain this information from sources such as Credit Karma.
Another growing auto loan fraud trend involves scammers who create a phony identity rather than steal an existing one, a crime known as synthetic fraud.
“Synthetic fraudsters look like real people with great credit scores and well-established employment, which makes it very difficult for dealership personnel to identify,” TransUnion Senior Vice President Satyan Merchant said in a July statement.
Merchant in January said the synthetic fraud rate in auto lending had grown nearly 30 percent from the first quarter of 2021.
“Incidence of synthetic fraud in auto lending has grown faster than any other financial sector as we emerge from the pandemic,” he said in a statement.
Another problem involves what Point Predictive calls “fraud for car,” when consumers who need transportation lie about their income or employment to qualify for a car loan. Such lies can also arise in “fraud for profit” cases, according to the company.
Point Predictive had previously focused on serving auto lenders, who would alert dealerships or request stipulations after a questionable application had been submitted. With the new BorrowerCheck system, dealerships can tap the same Point Predictive database and perform such due diligence themselves. A dealership could recognize which additional documentation, such as a customer’s employment, would be necessary before even submitting an application to the lender, Point Predictive said.
“Dealers play a vital role on the front line, helping to validate for lenders whether a prospective borrower is being truthful on their applications for financing,” Point Predictive CEO Tim Grace said in a statement.
BorrowerCheck also gives dealerships a verification system that even Point Predictive’s lenders didn’t have before, according to McKenna. Given the issue with traditional identity validation questions, the company opted to focus on a consumer’s phone number, which he said seemed more difficult to fake.
Point Predictive’s software checks records for signs the phone number provided by a loan applicant is associated with the person they’re purporting to be. If the phone number seems legitimate, Point Predictive lets the dealership send a one-time passcode to that phone and request confirmation of it by the applicant.
In all but one instance of identity theft tracked by Point Predictive, the fraudster used a different phone number than the one belonging to the person they’re claiming to be.
McKenna said dealerships running these initial checks could speed the process for legitimate customers. The one-time passcode is faster than verification questionnaires, he said. Dealerships can recognize applicants who will need additional documentation and request it before even submitting the loan rather than waiting for a lender to demand it later, he said.
The screening also permits retailers to send better deals to lenders, reducing the risk of buying back early defaults and preserving the relationship with lenders who judge dealers on loan quality. (Point Predictive noted in its BorrowerCheck announcement Aug. 30 that it sells lenders the DealerCheck service to conduct such monitoring.)
“Our customers’ experience is vitally important to us,” Ryan Morris, vice president of corporate finance at Southern California-based Mossy Automotive Group, said in a statement. “We believe limiting fraud risk exposure and protecting our dealership reputation while prioritizing our customers’ experience is the key to our continued success. Lowering our risk of buy-back demands will also increase our bottom line.”
CreditMiner and TransUnion already had partnered on one industry facing tool released this year. The Synthetic Fraud Score announced in January rates the probability of an applicant being a scammer on a scale of 100-1,000.
Luna said a monthlong pilot program found 3 percent of prospective borrowers receiving soft credit checks through CreditMiner had synthetic fraud scores exceeding 500, which meant a more than 50 percent chance of a scam. (Some legitimate customer behavior, such as a recent divorce or a move, can cause false positives, he said.)
While 3 percent doesn’t sound like much, “it really is,” Luna said.
If the score is high enough, CreditMiner recommends a dealership use its new Identify software to scrutinize the person and their ID more closely, Luna said.
The new software checks the authenticity of the driver’s license, compares the personal information on it to TransUnion records and examines whether the ID photo matches the selfie taken by a customer as part of the system. It also checks the phone to see if the device has been tied to suspicious activity.
“It’s not just the individual,” Luna said. “It’s that phone.”
CreditMiner has seen interest from large dealership groups in adopting the product and said one planned to just skip the synthetic score phase and just run the Identify check on every consumer.