Leila Shanoff's mandate a year ago when she was hired at Heartland Financial USA was to help the company make quicker small business loan decisions, so it could boost revenues derived from the segment.
The senior vice president and manager of Heartland's small business loan center immediately provided what she was confident would be the answer.
"We need a platform that houses LiquidCredit," Shanoff recalls saying, referring to Fair Isaac's credit scoring solution. "Either you do your own custom scorecards or you use Fair Isaac scorecards, which is what I've used for 33 years. We've housed [the software] on a lot of different platforms and it's worked well. If I'm going to take on a portfolio like this, I want to compare it to national statistics, and I knew FICO had that."
Heartland, a $5 billion-asset holding company of nine banks based in Dubuque, Iowa, was set to begin deploying the solution at press time.
Competing in the small business loan segment can sometimes be difficult for smaller banks, particularly if the larger competition aims their bigger budgets toward deploying automated scoring at high-volume business.
"A bank of this size typically shies away from small business," Shanoff explains. "If it had to underwrite a $250,000 or smaller loan without such a platform, it could take weeks. But if I could put it on a scorecard, I could get a decision in 24 hours. So this is the only type of platform that would help us compete against the Chases and the Wells Fargos, as well as credit unions that turn these deals around. The only way to compete is to bring in a scorecard that can help me get a decision within 24 hours. I have to get that customer off the market, because of the competition."
FICO's LiquidCredit analyzes companies' financial information, which it gets from Heartland's commercial lending frontend, SunGard's Ambit and Custom Credit Systems' CustomLender. After receiving data from as many sources as possible, including companies' loan applications, tax returns if they have any, and data from commercial credit bureaus like Dun & Bradstreet, LiquidCredit crunches the numbers and produces a score based on national loan performance statistics and FICO analytics that can enable a bank to make a yes/no decision on a business loan. Sources with more thorough information are weighted more heavily in scoring, Shanoff says.
Shanoff is primarily applying a score-assisted model, in which she incorporates the score LiquidCredit returns as one element in overall decision-making. "So I'll do some judgmental lending in addition," she says. "Maybe up to a certain dollar amount, I'll use the score only. But if you talk to me in one year, I might say differently. This is a bank that never did small business loans, so I had no baseline; I have to start from scratch. And I wouldn't do that by saying the score's the only answer. But I'll still be able to make a decision within 24 hours."
FICO has a startup scorecard to help banks that have little knowledge of a company. LiquidCredit will compare the company to others of the same standard industry classification, or SIC code, meaning essentially firms of similar type and size, across national statistics. The data can be pulled from Dun & Bradstreet.
Drawing profits from low-dollar small business loans can require scale, as well as expertise.
"For a small bank it's probably more important to get a couple $10 million deals on the books rather than 100 $250,000 deals," Shanoff explains. "Whereas for Wells, those $250,000 deals are its bread and butter: They could put thousands of those on in a month and put it through a scorecard. That's similar to what I'm proposing here."
