Fair Isaac Tweaks Credit Score Formula

Fair Isaac FIC is reworking the formula for its credit score, which is widely used by banks, credit card companies, mortgage brokers, landlords and even insurers to determine whether consumers are a good risk.

The company says the changes, which take effect in September, will improve the accuracy of its scores by between 5% and 15%, particularly among borrowers with a history of missing payments as well as immigrants, young people and others with "thin" credit files.

Fair Isaac uses credit histories from the three major credit-reporting agencies, Equifax, Experian and TransUnion, to rank consumers on a scale from 300 to 850. Consumers with higher scores generally qualify for lower interest rates, while those with lower scores pay higher rates or may not qualify for a loan at all.

The company said the scale will remain the same once the changes are made.

The changes come amid sharp increases in the number of borrowers with low credit scores who are defaulting on their mortgages. Foreclosures were up 62% in April from a year earlier, according to RealtyTrac.

Fair Isaac says it tweaks its model periodically; while the upcoming reformulation is more extensive than some others, it is unrelated to the subprime mortgage meltdown.

"This refresh of our model is part of business as usual," spokesman Craig Watts says. "Since we introduced it in 1989, we have continually redeveloped the scoring model. Research and design has been working on it for months."

Despite the expected improvement in accuracy, Watts says most consumers will see little change in their numerical scores. He said the new formula could result in slightly better scores for some consumers and slightly worse for others.

The main purpose of the update is to "help lenders keep pace with the constantly changing dynamics of consumer credit behavior," says Watts. "For lenders, slight differences can be significant."

Fair Isaac expects the reformulation to help lenders weed out the bad risk from the good. "What we're really interested in is the gray zone," says Watts. "We want to focus on what's in between the obviously good and obviously bad risks."