The decision to extend credit to a borrower is fundamentally about trust. In credit markets, firms like Dun and Bradstreet help lenders determine who is trustworthy and who is not by examining the past behavior of potential borrowers. By instilling bankers with a sense of trust in borrowers whom they did not know personally, credit raters conceivably expand access to lending. But how true is the old adage that the best predictor of future behavior is past behavior? Are credit ratings accurate predictors of loan default?
Bruce Carruthers of Northwestern University previously took up this question in historical research on Chicago firms circa 1879. He found that the published credit ratings of the Dun Company were poor predictors of eventual loan default. With funding from the Russell Sage Foundation, Carruthers will expand his study to include more Chicago firms, as well as firms in the same era from Boston; Peoria, Illinois; and Springfield, Massachusetts. Adding more businesses will increase the statistical power of his data, since loan default was a rare occurrence in his original data set. Chicago’s economy was accelerating after the Great Fire of 1871, meaning that business failure was particularly rare. Adding Boston to his study will allow Carruthers to compare Chicago to a metropolis that had a slower rate of growth and therefore more firms that were at a high risk of loan default. Data from Peoria and Springfield will expand Carruthers inquiry to smaller communities where personal trust networks may be more important in determining trustworthiness than are credit ratings.