Is There a 'Boom Bias' in Agency Ratings?
Forthcoming in Review of Finance; published by Oxford University Press.
59 Pages Posted: 5 Oct 2013 Last revised: 30 May 2015
Date Written: May 1, 2014
Theory predicts rating agencies' incentive conflicts to be stronger in boom periods, thereby leading to biased ratings and a reduced level of rating quality. We empirically investigate this prediction using a large data set of almost 10,000 U.S. corporate bonds, publicly issued between 1990 and 2007. Our main findings are twofold: First, initial ratings appear to be overly optimistic during boom periods when compared to initial bond spread levels or 'incentive-free' benchmark ratings. Second, boom bond ratings are more heavily downgraded and perform poorly in predicting defaults from an ex post perspective. In several robustness checks we show that the observed 'boom bias' is not resulting from changes in credit-worthiness, adjustments in rating standards over time, competitive pressure, or investor demand, but rather from rating agencies' incentive conflicts.
Keywords: Credit ratings, credit risk, conflicts of interest, yield spreads
JEL Classification: G01, G12, G24
Suggested Citation: Suggested Citation