Bankers expect a lot when they commit to a new technology. They want the system to improve customer service, generate greater revenue opportunities, and pay for itself, if not signiWcantly reduce expenses. Few applications live up to that promise, at least right away. But Chemical Banking Corp. said the system it began using last year to rate the creditworthiness of small-business borrowers has exceeded its expectations.
Last June, the New York-based bank began to use a statistical scoring technique to assess small-business loans of up to $100,000. The system, from Fair, Isaac & Co., a pioneer in developing credit scoring systems for consumer lending, has sped up Chemical's approval cycle, enabled lenders to spend more time prospecting for new customers, and contributed to a 44% increase in small-business loan volume. Executives say that more companies are qualifying for loans now than before, and credit quality is expected to remain high.
It's a dramatic improvement from the way most banks handled small- business loans 12 months ago, said Carmen G. Mastroianni Jr., senior vice president for small-business lending at the $188 billion-asset bank. It has revolutionized the way we are doing it. But it has the potential to really further revolutionize the small-business process.
The scorecard, which gauges the probability that a borrower will repay a loan, is based on data collected from 17 banks by San Rafael, Calif.-based Fair, Isaac and Robert Morris Associates, a trade group of commercial lenders. The list of pilot institutions includes Chemical, Chase Manhattan Corp., and Hibernia Corp.
In standardized categories of loans under a certain limit, automation is absolutely essential to doing the business proWtably, said Patricia McGinnis, a technology analyst at the Tower Group, Wellesley, Mass.
But while Ms. McGinnis said the technology is an important tool, she cautioned that it is not applicable
to all categories of small-business loans. Loans that require complex collateral requirements are not an easy Wt, for example. Another reason is that the scorecard predicts creditworthiness, not actual losses on loans that go bad. It predicts the probability of default, but it does not predict how much you may default, she said. It doesn't tell you how much damage you are going to do to the bank.
But for many small-business loans, there are clear beneWts. Chemical reports that the time it takes to make a small-business loan decision has been cut from two to three days to less than 24 hours. A year ago, Mr. Mastroianni said, 40 loan oYcers made small-business lending decisions. Now, just three are needed.
The remaining small-business lenders continue to evaluate loans of more than $100,000, about 20% of the volume handled by the division. With their newfound time, lenders are expected to better serve existing customers and look for new business.
As a result of rechanneling those loan oYcers to more of a business development role, our volume increased dramatically, said Mr. Mastroianni.
Both the dollar value and number of small-business loans has jumped more than 40%, he said.
The business has also grown because more companies are getting loans than in the past.
There's been a shift, said Mr. Mastroianni. In the past, we would tend to look at just the Wnancial numbers of the business, and we wanted to make sure the business numbers would support the loan.
In contrast, the Fair, Isaac system assigns greater weight to the Wnancial resources and credit history of the business owner.
So far, Chemical hasn't entirely removed the human element from small-business lending decisions; loan oYcers continue to give applications a quick once-over. But Mr. Mastroianni said the technology is working well enough that the bank will begin to rely exclusively on the credit score for loans under $100,000.
The early success of the system will also probably spur Chemical to develop an additional scorecard to evaluate loans up to $500,000 or $750,000.
That would have a dramatic eVect not only on my operations but the entire low end of our middle-market operation, he said.
But unlike the Fair, Isaac system, a Chemical scorecard would be based on proprietary data.
I would be much more comfortable scoring the larger loans with my own card, because I know everything that's in it, said Mr. Mastroianni.
When Chemical's pending merger with Chase is complete, he anticipates that the combined bank will have a large enough data base to produce a statistically valid scorecard for larger loans.
The bank, he said, would develop the system in tandem with a credit scoring vendor, perhaps Fair, Isaac.
The decision to build such a decision will be made this year, said Mr. Mastroianni.
This has served as the catalyst for us to really automate the small-business loan process to a degree that we hadn't been able to before the scorecard, he said.
Chemical has simpliWed the application and further automated the documentation process, for example.
Mr. Mastroianni also foresees a day when loan oYcers will be able to visit prospects in the Weld with laptops equipped with the scorecard. After the Wnancial information is entered into the computer, the lender could give customers a loan decision on the spot.
If (the loan) is approved, we can prepare the documentation as well and have that printed out at the customer's oYce, he said. And by the end of the visit - through automation and credit scoring - we could have the loan approved and booked, and documentation signed.
Mr. Mastroianni said switching to the new technology was relatively painless.
The staV quickly learned the new front-end processing system supplied by Fair, Isaac, he said.
But some employees initially feared that automation would result in major layoVs. Mr. Mastroianni eased those worries by noting that Chemical was committed to expanding its small-business lending.
The bank has not said how the small-business lending staV would be aVected by the merger with Chase.
But Mr. Mastroianni expects the combined bank to devote more resources to building small-business lending.
He added, Our branch folks and sales folks that work within the branch system seem to have been energized by the fact that we are able to (make decisions on) these loans much more quickly than in the past.