Bank reports 60 percent more loan approvals by using alternative data

'It's not sacrificing or compromising, what people would think is subprime or near-prime at all.'

When a group of industry experts met this week to discuss using alternative data in credit underwriting, Peter Renton asked the panel how alt data could help expand the “credit box” or bring more financial inclusion to lending.

Marc Butterfield, SVP of Innovation and Disruption at the First National Bank of Omaha, said that the results were significant when FNBO ran tests. For example, he said using alt data helped approve 60 percent more applications than the old model.

“We did some head-to-head tests of our traditional model versus our model plus alternative data for that conservative unsecured consumer lending product,” Butterfield said.

“The simple example I will use is this: there were 10 borrowers that the bank declined. And when we added in those additional data points, we would have approved and did approve six of those 10.”

Butterfield, representing a 165-year-old bank that has been handing out consumer credit since before the first credit card companies, said his data showed creditworthy customers were being overlooked. There was a significant inclusionary difference when you improved lending through including alternative data, he said.

“So far, this is now going on a two-year-old test and the performance of that data is as if we were to approve them with our own model, in terms of account performance. So, it’s not sacrificing or compromising, what people would think is subprime or near-prime at all.”

Several forms of alternative data

Butterfield’s pro data response came toward the end of Tuesday’s webinar, after comments from Luis Moneda, data science manager at NuBank, and Chris McKay, chief credit officer of OppFi.

Moneda said alternative data could come in the form of texting and mobile application data.

When it came to alternative data for underwriting, McKay said “I don’t think we can operate our business without it.”

“Assuming your goal is to lend as much as you can within certain thresholds of risk, then by doing that, you’re going to expand the box and that’s exactly what we’re looking for, we find that every time we add additional sources, it’s allowing us to create that.”

In closing out the webinar, Bence Jendruszak, COO of SEON, summed it up: Using alternative data to improve approvals helps the bottom line.

SEON is a fraud prevention startup that establishes customer digital footprints to take out fake applications.

Of course, knowing how to weed out bad actors is important, but Jendruszak said it’s important to think in positive terms: alt data can improve customer approvals, it’s not just for eliminating fraud.

“I definitely believe that it’s a good thing to be behind right I mean, I touched upon this point when you were asking about fraud and the different use cases, and I do believe that it’s all about you know, upping your bottom line and actually accepting more good clients rather than always thinking about the dishonest actors out there. At the end of the day, it’s the right way to go about it to think about the bottom line.”

Intensely energetic news reporter who finds inspiration chasing breaking news stories and covering the collision of Silicon Valley and Wall Street. Experienced in news writing, digital media production for online and print publications, newspapers, and magazines, as well as copy editing and press release revisions.

He studied history at the University of Delaware and wrote for the Review.

Email me at kevin@lendit.com with story ideas, questions, or just to say hello.