Westlake Financial Services is testing pilot programs that use alternative data to digitally verify consumer credit, which the company hopes will lessen the amount of fraud incurred,Chris Urban, vice president of risk management at Westlake, told Auto Finance News.
“It doesn’t necessarily change the speed of the approval, for us it’s more about improving the predictiveness of our approval,” Chris Urban, vice president of risk management at Westlake, told Auto Finance News. “While our approvals are automated there are still verifications to be done and our focus now is to do as much of that verification automatically when the dealer sees our approval, before they send it in.”
The alternative data used includes utility and cell phone bills as well as property ownership and address history gathered by vendors such as Equifax, Experian, Lexis Nexis and others, Urban said. While there are several companies such as Hyundai Capital America, that are testing similar programs, Urban believes these systems could one day be fully automated to the point where no human has to be involved with the individual deal.
“Potentially, it could even be more reliable than human verification and that’s a big win for us,” Urban said while also adding that humans will never be completely devoid of the process.
“It’s way down the line when decisions get booked, approved, and verified without a human ever looking at it,” Urban said. “Even then, let’s say we did do that, we would still want to look at a very large sample of individual deals we’ve booked and look at them with a human eye and say ‘are we missing anything?’ We’ll never just sit back and say ‘The computer is perfect.’”
In the short term, these pilot programs can still increase confidence in the quality of the orginations, but Urban says it will be a gradual process.
“If it turns out a customer doesn’t make as much as they say they would, ideally I don’t want to find that out several days later when the deal’s already in house and we’re talking to the employers, I want to find that out right away using additional data sources,” Urban added. “The more uncertainty you take away from the back and calculate in the front end, the better.”