• Erik Berwart

Student thesis: Phd


In this thesis, we study if credit rating agencies (CRAs) are capable, through their rating process, of discover information that it is valued by the market. Additionally, we investigate if CRAs are able to propagate their findings to the market. if Specifically, we study the differences between issuer-paid and investor-paid credit rating agencies, and how those differences shape the characteristics of their credit ratings and ultimately, if investors can profit from credit rating announcements.For our research we use a large dataset of rating announcements from 1997 to 2012, which includes information of four credit rating agencies (CRAs), Egan-Jones Ratings Company (EJR), Fitch, Moody's and Standard and Poor's, which representing investor-paid and issuer-paid CRAs. This allows us to compare these two kind of agencies and its ratings.In the first essay we study what variables explain the rating coverage of an investor-paid credit rating agency. We show that probability of being covered by EJR is positively related with the size of the firm, the level of institutional ownership of the firm, stock analysts and issuer-paid CRAs level of coverage, while it is negatively related to the firm's corporate governance. We found that the likelihood of being covered by EJR augments after regulatory changes and most interestingly, since EJR received the NRSRO certification.In the second essay we compare the timeliness of rating changes produced by EJR and the issuer-paid CRAs representatives. We found that the lead effect of investor-paid over issuer-paid CRAs has weakened in recent years, while Granger causality is bidirectional and therefore a lead-lag relationship cannot be established. Finally, stock prices manifest statistically significant abnormal reactions to downgrades of all agencies; however, abnormal negative returns are significantly higher for EJR. Our results support the hypothesis that issuer-paid agencies improve the quality and timeliness of their ratings when they see their market power threatened by tighter regulations. Nevertheless, event studies illustrate that markets still price stocks under the assumption that investor-paid rating actions carry superior information.Finally, our third essay found that purchasing (selling short) stocks with positive (negative) rating announcements generates portfolios with positive annual abnormal returns when investors react immediately to rating announcements. Returns are higher for stronger announcements (i.e. rating changes over rating outlooks) and for an investor-paid agency rather than an issuer-paid agency. When we introduced transaction costs, only the investor-paid agencies' announcements lead to positive abnormal returns. Additionally, when we included a delay in the reaction of investors to rating announcements, all positive abnormal returns net of transaction costs disappeared. Finally, our results suggests that the differences between investor-paid and issuer-paid agencies are based on their dissimilar business models rather than their regulatory status.
Date of Award31 Dec 2015
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorMassimo Guidolin (Supervisor)


  • Credit rating agencies, Issuer-paid agencies, Investor-paid agencies
  • Timeliness, Conflict of interest, Portfolio strategies

Cite this