One financial technology company CEO says AI and machine learning have opened new doors for credit lending. 

Propel Holdings, an online financial technology service, recently reported a 47 per cent jump in the company’s first-quarter sales, a result CEO Clive Kinross attributes to AI and machine learning that is able to change the game of traditional credit scoring.

“Our AI and our machine learning looks at thousands and thousands of variables and over 5,000 variables in each customer as compared to traditional credit scores which are just not useful in scoring our consumers,” he told BNN Bloomberg during an interview on Monday. 

“As a result of that we’re able to find more and more consumers because of our AI and machine learning without necessarily taking on more risk,” he explained. 

Propel Holdings provides credit to “underserved consumers across the U.S. and Canada,” Kinross said. “We’ve been doing that for the last 12 and a half years or so on our proprietary AI technology platform.”

He mentioned that North America comprises about 70 million consumers who cannot access credit from mainstream banks and credit unions. 

“We have bank partners who have a deep desire to be able to provide credit to these types of consumers,” Kinross said. “They don’t have the expertise nor the skills so they partner with us and together we are able to create leading programs to provide access to credit to these consumers in products that are vastly superior to products than what is otherwise available on the market.”

According to Kinross, Propel Holdings’ AI and machine learning evaluates over 5,000 variables in each customer “as compared to traditional credit scores which are just not useful in scoring our consumers.” 

This leads to more consumers without taking on more risk, he says.

Kinross also mentioned that the interest rate environment can reduce margins. 

“We fund our loans through debt,” he said. “Debt costs have gone up. When we went public our cost of debt was around nine per cent. Today, just because of the rising interest rates, it’s 13.5 half per cent.”

He explained that the incremental four and a half per cent in debt costs act as a “headwind.” 

“We think we’re at the top of the credit cycle as those rates start to come down. Obviously, the difference will go to our bottom line. But from our perspective, our rates are fixed and if anything we have a graduation program where as consumers prove their pay over time we lower their cost of credit.” 

Kinross added that margins for the business have increased “materially” over recent years. 

“As an example, in Q1 of this year, we increased our net profit by 77 per cent,” he explained, pointing to US$13.1 million in adjusted earnings which are “largely driven by AI and machine learning.” 

Kinross acknowledged that traditional credit scoring could still be useful for “evaluating prime and super prime consumers” but that leaves out a large portion of consumers.

“We need to look at a whole host of other scores that are predictive as to whether the consumers are on a path to recovery,” he said. “We like to finance them when they’re at that point in the cycle.”

Watch the rest of Kinross’s interview with BNN Bloomberg,