In a March working paper published by the Federal Deposit Insurance Corp., FDIC Associate Director of Insurance & Research Robert De-Young, Deputy Director Dennis Glennon at the Office of the Comptroller, and Bryant University finance professor Peter Nigro set out to compare the default rates for loans backed by banks that used credit scores with those that did not. They found that among banks that didn't use credit scores to evaluate loan applications, the likelihood of default rose as the distance between the location of the borrower's business and the lending bank increased. A borrower more than 50 miles away from the lender was 22% more likely to default than one less than 25 miles from the lender.
But distance wasn't a factor in default rates at banks that used credit scoring. That's an important finding, as the average distance between borrowers and lenders using credit scores was 142 miles in 2001, up from just 12 miles in 1995.
While using a borrower's credit score improves a lender's ability to gauge risk, it has not actually led to lower default rates. In fact, the default rate is 23% higher for loans in which banks use credit scoring vs. those that are not scored. That's because banks, particularly the largest lenders, are relying more on credit scores than on visits to an applicant's place of business, significantly cutting their costs.
"This raises the profit margin on small business loans, allowing lenders to still make money on those additional, riskier borrowers," says DeYoung. Indeed, the use of credit scoring may have cut banks' costs enough that they're now comfortable taking on more risk, even if it means accepting higher default rates on some small business loans.
The problem is exacerbated when lenders who use credit scoring repackage the loans and try to resell them in the financial markets. Reselling the loans gives the bank the cash to make another round of loans to potentially riskier borrowers--based largely on their credit scores. By James Mehring