HOME Recent Events RMI in the News ALUMNI
  Issue 15 | Archive May 2013

Time-varying Rating Standards and the Distorted Incentives of Credit Rating Agencies

A paper by Tao Wang (National University of Singapore)

PhD candidate Tao Wang from the Department of Finance at the National University of Singapore, who is supervised by RMI’s Director, Prof. Duan Jin-Chuan, tests whether reputation concerns could always discipline credit rating agencies in his working paper.

The credit rating agencies (CRAs, henceforth) play an important role as financial intermediations in the modern financial system. However, the CRAs have often been criticized for revealing biased information to the investors. At the heart of these criticisms lies the issuer-pay business model adopted by major CRAs, which means issuers of different securities pay for the credit ratings rather than investors. It has raised concerns that the incentives of the CRAs are distorted, which leads them to issue more favorable ratings to attract/retain clients. In response to the criticisms for their conflicts of interest, the CRAs always defend themselves with a counterargument of having an incentive to build and protect their reputations for being independent and objective. The author thus investigates whether such reputation concerns could serve as a self-disciplining mechanism for rating agencies by studying the time-series pattern of rating standards.

The author argues that credit ratings are higher due to lax rating standards after controlling for firm level characteristics when the corporate bond credit-yield curve is steep. Corporate bond credit-yield curve refers to the chart of average corporate bond yield with respect to credit ratings. When the slope of such curve is high, bond market investors require much higher return for risky projects than the safe ones. Bond issuance costs would thus increase significantly if the issuing company were assigned a rating lower than what is expected. As a consequence, the likelihood that companies with unsatisfactory ratings cancelling their potential bond issuance becomes much higher in such case and the CRAs could lose a large amount of business if they still conducted ratings rigorously. In order to secure their income, CRAs are more likely to inflate the ratings in such conditions but become relatively more stringent otherwise if the reputation concerns are not always powerful enough. Therefore, a direct testable hypothesis is that credit ratings tend to be higher when the corporate bond credit-yield curve is steep after controlling firm characteristics.

To test such a hypothesis, the author conducts empirical analysis using a sample of S&P’s corporate credit ratings from 1985 to 2011. Regression analysis confirms the hypothesis that credit ratings are higher given a steep corporate bond credit-yield curve after controlling firm level characteristics as well as industry or firm fixed effects. Such time-varying rating standards suggest that CRAs’ incentive problems are more serious during periods when bond market investors are more risk-averse. As a comparison, the author also conducts the same analysis using the probabilities of default (PDs, henceforth) released by the Credit Research Initiative (CRI, henceforth) of Risk Management Institute (RMI, henceforth) at National University of Singapore. The RMI-CRI system adopts an open platform, making technical details public, and the access to the results is freely available. The regression results based on such objectively constructed PDs are totally different from those based on credit ratings from for-profit credit rating agencies and serve as an instrument for comparison.

The author also argues that the relationship between the corporate bond credit-yield curve and rating standards is asymmetric. Credit ratings are inflated when the credit-yield curve is steep but rating deflation does not take place given a flat credit-yield curve. The empirical analysis is consistent with the asymmetric hypothesis and provides little evidence for rating deflation. The author also finds that ratings for complex firms are more likely to be inflated when the slope of corporate bond credit-yield curve is high. The intuition behind this phenomenon is that CRAs could attribute the inflated ratings to the complex business that firms are engaged in rather than their distorted incentives. And the reputation loss would not be too large in case of such complex firms with inflated ratings default. In addition, the author also finds that CRAs are more likely to inflate ratings for companies with higher credit risk. This is because the slope of corporate bond credit-yield curve is usually higher for bonds with low credit ratings. Therefore, risky firms are more concerned of their ratings and the additional issuance cost will be much higher for such firms if the ratings assigned to them are lower than expectations. And CRAs would have stronger incentives to inflate ratings for such companies.

In conclusion, this study sheds light on the discussion regarding CRAs' incentives to fairly reveal information under the issuer-pay model. It also provides justifications for the reform of the rating industry through either the regulatory route or a more fundamental approach of overhauling the for-profit credit rating business model by adding a public-good alternative as described in Duan and van Laere (2012).

Back To Newsletter

Published quarterly by Risk Management Institute, NUS
Editor: Ivy Wang (rmiwy@nus.edu.sg)