Determinants and Sensitivity of Credit Ratings: A Panel Data Analysis

Abstract

Using panel ordinal logit model from 2009 to 2019, this study analyzed the determinants of corporate bond credit ratings. Although credit rating agencies consider both market and financial factors to determine rating (Agency Rating, AR), most literature only considers the effect of financial factors on AR. This study attempts to fill this gap by adding market risk measure and analyzing the effect on bond-implied rating (BIR) determination. We also analyze the sensitivity difference between AR and BIR. Our finding reveals that both AR and BIR are significantly improved when volatilities in sales and operating profit increase. Regarding rating inflation (mismatches), increases in assets further improve AR, while net exports and industrial activity lower BIR, leading to rating inflation. The increase in M2 (broad measure of money supply) worsens AR while improving BIR, leading to rating deflation. Finally, the effects of large enterprises, recession, and construction industries on the rating inflation is analyzed through dummy variables. The results revealed that, unlike in previous studies, the dummy variable of large enterprises was not significant. The relative worsening of BIR during the recession resulted in rating inflation. Finally, the stigma effect on construction industry was observed, which shows significant rating inflation.