Credit scoring has drawn skeptics and critics, but lenders continue to find new uses for it - and companies that create decision tools continue to offer new products for lenders' changing needs. Credit scores are numerical ratings that lenders use to forecast a person's credit performance. Although scores historically have been used to assess the likelihood that a person will default, lenders today are finding a wider range of uses for them.

"Risk management is a money-generating department now, not just a loss- control department," said Lurdis Abruscato, editor of Credit Risk Management Report, an industry publication.

Of course, other risk management systems exist - such as neural networks, which many industry experts complain are difficult to use. But credit scoring seems by far the most popular tool in a wide range of industries.

"The bottom line is that the biggest credit issuers - GMAC, NationsBank, Banc One, and Bank of America - all go back to a proven product . . . more traditional scoring techniques," said Rawdy Shediac, senior vice president of application screening products for Fair, Isaac and Co. in San Francisco. "They have been more sophisticated in how to use our tools, and we have become more sophisticated in how to develop new tools."

One of the newest applications of credit scoring is among bankers and credit card issuers who use it to attract and retain customers. Instead of using credit scoring simply to disqualify customers who do not meet criteria, lenders can use it to help with marketing - for different products, rates, fees, and terms.

Borrowers who do not qualify for top-tier credit at the lowest cost still may be eligible for some level of credit, which makes it available to more people, industry experts say.

In addition, lenders are using new credit scoring models to determine how much revenue customers are likely to generate and the likelihood that certain customers will seek credit elsewhere. Fair Isaac's Mr. Shediac predicted that scoring the creditworthiness of small businesses and marketing through preapproved mailings to those that qualify is the next big thing on the credit scoring horizon.

While credit scoring has been used for years by credit card issuers and other lenders, it just now is beginning to take hold in the mortgage industry. Much of the impetus for this came last summer when Freddie Mac, the Federal Home Loan Mortgage Corp., issued an industry letter lauding the benefits of scoring.

The agency does not currently require lenders to apply credit scoring to their loan underwriting - and it probably will not have to, said Bruce Wood, director of Freddie Mac's mortgage credit policy. Mortgage lenders are likely to see the benefits of credit scoring on their own, he said. In the past, Freddie Mac issued guidelines on the income-to-debt ratios that mortgage lenders should enforce. But before credit scoring, there was no benchmark for judging a borrower's credit reputation, Mr. Wood said. Lenders are finding new uses for credit scoring, especially in connection with loan servicing. "There is a lot of interest in being able to protect portfolios from refinancing," he said.

However, some in the industry are skeptical of the advantages that credit scoring may offer. Kathy Schroeder, vice president of corporate underwriting at PMI Mortgage Co. in San Francisco, is one doubter.

PMI markets its own risk scoring model, which it developed in 1987 and has commercially licensed to customers since 1992. Unlike credit scoring, in which a numerical score is based on a borrower's credit history, PMI's model weighs several factors.

"That one number is an important aspect, but it shouldn't be the sole determinant for whether or not a loan should be made," said Ms. Schroeder. "We look at other factors that are important in overall risk." Ms. Moore is a freelance writer based in Los Angeles.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.