ZestFinance developed and recently launched a lender tool that uses machine learning to originate loans for thin-file borrowers, and the company is already working with one of the “Big Three” captives to implement this system for its auto originations, said Douglas Merrill, founder and chief executive of the company.
The platform applies algorithms to alternative data using a process called machine learning, which is a type of artificial intelligence that allows the technology to learn on its own without manual input, Merrill said.
He developed the platform using methods he learned from his former position as the chief information officer at Google. The unnamed lender wants to engage in the consumer’s first vehicle purchase in order to build brand loyalty, and Zest’s platform aims to make lending to those millennials less risky, Merrill told Auto Finance News.
“[The captive’s] view is that if your first car is their brand, your last car will be their brand,” he said. “So they are pretty focused on ‘How do I make sure the first interaction is a good one?’”
At first, the captive didn’t want to extend originations to consumers with Fico scores below 720, but through machine learning, losses below that range have decreased 23%, Merrill said. The company is confident its algorithmic platform could reduce losses and allow lenders to more safely lend into the Fico score range of the low 600s, he added.
The platform also allows lenders to better monitor disparate impact and change the model to quickly limit unintentional discrimination of protected classes.
“We took BISG and built on top of it to expand it,” said Merrill, a former researcher at the RAND Corp., which developed the method now used by the Consumer Financial Protection Bureau to identify disparate impact. “We should speak the regulator’s language, and because it’s so easy to do [on Zest], you can protect yourself even from accidental discrimination.”