OakNorth: How AI Can Fuel Banks’ Return To SMB Lending

Small businesses (SMBs) left behind by their banks, thanks to tightening risk and capital requirements post-financial crisis, certainly left a gap in the SMB lending market that alternative and marketplace lenders rushed to fill. In the U.K., support for FinTech, as well as challenger banks, to boost competition in the small business lending market is driven by the oft-repeated conundrum for this industry: small businesses are more complex than a consumer, yet not as lucrative as a large enterprise, and are, therefore, unattractive borrowers to traditional, large lenders.

With a growing community of alternative finance (AltFin) players, banks have taken to either letting AltFin handle filling the small business lending gap or partnering with and acquiring those FinTech firms to retain their SMB customers. Yet, a gap remains. While the Goldilocks Conundrum offers a simple explanation as to why, a deeper dive into the issue of small business bank lending reveals a far more complex issue.

According to OakNorth Analytical Intelligence, both large banks and alternative FinTech firms are struggling to overcome the challenges of SMB lending, and it often comes down to data. Amir Nooriala, COO of OakNorth, said that to underwrite an SMB loan, financiers have to take in and analyze unstructured data.

“In micro-lending, P2P and retail, it’s structured data,” he told PYMNTS. “It’s easy to source; it’s rule-based. I look at your credit score and your bank statement and card statements, and the machine gives a ‘yes’ or ‘no.'”

The SMB space, however, offers lenders troves of unstructured data in various formats. Take, for instance, the example of a hotel that needs a $5 million loan to expand. That business may have financial data recorded on Excel spreadsheets in a structure and format that a bank has never seen before. Data might be kept in physical bank branches. Information must be translated if the company is outside of an English-speaking market, which could lead to errors or miscommunication. Yet, all this information is important to underwriting a loan.

“It’s going to be very bespoke to them,” Nooriala continued. “You need advanced technology to translate that data, to take unstructured data and structure it, and give meaning to it.”

On top of this challenge, he added, banks are challenged to analyze tens of thousands of different market scenarios and more abstract data points. Again, in the example of the hotel, it could be data points like average cost per room, average occupancy rate and the fact that Brexit has caused the value of the pound to decline, causing vacationers to “stay-cation” in the U.K. and, therefore, giving more business to local hotels.

“You need artificial intelligence [AI] to run through the scenarios,” said Nooriala. “At the corporate level, that’s maybe 64,000 scenarios — you can only have a computer do that, and you need structured data to do that.”

For large financial institutions (FIs), financing a multinational conglomerate with a loan worth millions of dollars makes this analytics process worth it. For SMBs, however, margins are too thin to have teams of experts aggregating, standardizing and analyzing this data. This, Nooriala noted, is the crux of the SMB lending gap.

Competitive pressure from FinTech firms, encouragement from policymakers and a broader understanding of that gap have led the big banks to step back into the market, either by collaborating with those FinTech firms or exploring how to increase their SMB loan books independently. Nooriala said traditional FIs have to make a decision about whether they want to allow FinTech firms to take the reins, and, if not, what the best course of action is to dig their heels deeper into small business lending.

However, the big banks’ pullback from SMB lending isn’t the only residual impact that the global financial crisis had on the market, both in the U.K. and elsewhere. According to Nooriala, that crisis has made traditional FIs open to trying to things.

“When there is a crisis shock to the market, it wakes people up,” he said — adding that, for some, it’s led to fear and protectionist action, while, for others, it’s forced FIs to examine how to stand out in a crowded market. “Banks are looking at what they currently have and trying to figure out what they need to acquire and upgrade. Are they willing to accept that there are third parties out there that will do something better than they ever could?”

For the institutions that aren’t willing to accept that, many are turning to those third parties as collaborators. For companies like OakNorth, which provides FIs with the technology to standardize unstructured data and automate underwriting processes using AI, this willingness in the traditional banking sector to explore the options has certainly been a positive impact from the global financial crisis.

“What the financial shock events have done is … wake [FIs] up, as opposed to forcing them in a single direction,” Nooriala said. “It’s definitely opened doors. People want to try something new.”