When determining how to address the shortage of small business funding options, Lendr founder Tim Roach visited a past success and gave it a twist.
For the past 16 years Mr. Roach has also worked as a managing partner at Oak Street Trading, a Chicago-based proprietary trading firm. He said Oak Street Trading focuses on providing advanced trading technology.
“Better technology, better systems and a unique way of looking at risk.”
Along with four partners, Mr. Roach founded Lendr in 2011, convinced they could employ better models when measuring risk. Based in New York City since mid-2016, Lendr quickly gained momentum by leveraging their independent sales organization network as they formed direct channels.
The trading world is well-ahead of lending in terms of technology, Mr. Roach said. Portfolio ratios can quickly be adjusted to reflect current market conditions and mitigate risk, for example. As he looked to apply trading lessons learned to lending he discovered how far behind lending technology actually was.
“In 2011 the technology was not there yet,” Mr.Roach admitted.
So in mid-2016 Lendr essentially blew up their legacy system. They hired a CTO and development team to create technology that separated them from the majority of marketplace lenders.
“The problem with most underwriting models banks use, is it’s a rear-facing model,” Mr. Roach explained in discussing how FICO scoring and other elements of many underwriting systems rely on past behaviors instead of predicting future ability. “The forward-facing model provides better decision making and better use of technology.”
Lendr offers client cash advances that can get approved within two hours and funded within a few days, Mr. Roach said. Repayment is made via a percentage of daily bank deposits until the entire amount is repaid. That allows companies that a rear-facing model might deny to access funds because Lendr’s system looks at current daily cash flow trends.
While the scoring system isn’t perfect, it’s trending in the right direction as Lendr’s technology team constantly improves it, Mr. Roach said. It is now at 95 per cent accuracy within 2.5 hours with a goal of 99.
And more data mean better decisions in the future, he added. Now armed with seven years’ worth of data, Lendr could look at older defaults to see who defaulted and why. Among the lessons were they had to look at each industry differently due to invoicing cycles and even tax implications. That takes us back to technology where machine learning and artificial intelligence allowed them to drill down into specific industries to understand the unique situations they face.
“We adjust models quarterly, sometimes faster,” Mr. Roach said. “We weren’t an agile company three or four years ago, but we’re getting to the point were we are.”
Like this article? Take a second to support us on Patreon!