“ Noise ” in Ratings : Not Entirely Random

“Noise” in Ratings: Not Entirely Random I investigate the informational accuracy as opposed to informational content of credit ratings over time. The study presents an empirical analysis of the classification errors. Results suggest that greater the credit risk in the economy, the higher is the error rate. Of all binary classifications tested, classification error for dichotomous class of investment vs. non-investment grade issuers is best explained by the economy-wide credit risk distribution. I conclude there is more “noise” in the assignment of ratings at exactly the time when there is more uncertainty regarding the credit risk of firms in the economy – i.e., during a credit crisis.