HOME Recent Events RMI in the News ALUMNI
  Issue 11 | Archive May 2012

A Public Good Approach to Credit Ratings – From Concept to Reality

A paper by Jin-Chuan Duan (National University of Singapore) and Elisabeth Van Laere (National University of Singapore)

In a forthcoming paper to be published in the Journal of Banking & Finance, RMI director Prof. Jin-Chuan Duan, together with RMI research fellow Dr. Elisabeth Van Laere, suggest a new approach to credit ratings, which is based on the non-profit undertaking by RMI and is premised upon the concept of credit ratings as a public good.

The paper starts with a critical analysis of the ongoing challenges facing the credit ratings industry. External credit ratings agencies (CRAs) have long been criticized for their opacity, questionable methodology, lack of accountability and conflicts of interest. Starting with the fall of Enron, Parmalat and the 2008-2009 financial crisis, the credibility of the CRAs has taken a serious hit. Once again the CRAs have been unable to effectively provide accurate information necessary for the proper functioning of the financial market. The atypical market structure of the industry seems to complicate the supply of meaningful and accurate ratings. The lack of competition, the for-profit motive of the CRAs and the inherent conflicts of interest have often been identified as the major sources of failure of rating agencies. The only defence of CRAs against these conflicts of interest and failures is the stake in their reputation.

The authors classify credit ratings into buy-side and sell-side, where the credit ratings issued by familiar CRAs such as Moody’s, S&P, Fitch fall in the sell-side category. It is the authors’ contention that the sell-side credit ratings market whose business model is based on the "issuer-pay" model needs to be reconfigured. The article continues with an overview of the developments and shortcomings in the regulations of the credit rating agencies. Since the 2008-2009 crisis, CRAs have been under enormous scrutiny and numerous changes and developments in the regulatory framework have occurred but without significant improvements. Changes have been focused both on closer supervision of the CRAs and reducing reliance on the external CRAs rather than overhauling the whole system or taking an alternate approach. The authors make a case that simply tinkering with the technical aspects of the existing business model of sell-side CRAs is insufficient and introduce the Credit Research Initiative as a promising alternative.

The Credit Research Initiative (CRI) of the RMI, conceptualized in March 2009 by Jin-Chuan Duan, is a constructive response to credit rating reform. The CRI is a non-profit undertaking premised upon the concept of credit ratings as a public good. Being a non-profit undertaking the CRI is meant to be non-proprietary and completely transparent. Operationally the CRI takes a two-pronged approach: (1) build up a comprehensive infrastructure for research, development and production of corporate default predictions, and (2) tap the worldwide credit research know-how through a “selective Wikipedia” style undertaking to develop organically evolving default prediction models.

In July 2010, the CRI began to release the daily updated probabilities of default (PDs) for over 17,000 exchange listed companies across 12 Asian economies. As of December 2011, the CRI releases daily updated PDs for over 28,000 active, exchange-listed firms in 30 economies in Asia (12), North America (2) and Europe (16). Including the delisted companies, the CRI covers around 50,000 firms in these 30 economies. The CRI operates on a proprietary database covering macroeconomic, financial and default related information for over 90,000 exchange-listed companies around the world. In addition to the 20 full-time staff, the CRI taps the global research pool by inviting external research teams to contribute to this initiative. This approach can be seen as a "selective Wikipedia" and will allow the implemented model to remain current, evolutionary and organic. Changes to the model are only made after thorough testing for operational stability and feasibility.

The current CRI model is a forward intensity model (reduced-form model) that allows default forecasts to be made for a range of horizons. The model is based on 12 input variables including common and firm specific risk factors. The key output is the release of daily updated PDs from a horizon of 1 month to up to 2 years for individual firms and various aggregates at industry and economy level. Providing this more nuanced credit information instead of a simple letter grade is in itself already a major step forward in the credit rating reform. Furthermore, the authors show that for a sample of companies that have a CRI PD and receive a rating from S&P, the CRI performs better than the for profit S&P ratings. The article also provides a critical analysis on the deficiencies of the CRI model and mentions a number of likely directions the CRI model might evolve to. Some of the likely future evolutions fall in the domain of further infrastructure development. For example, by end 2012, all exchange-listed firms around the globe should be covered. Other likely future directions will rely on the collective efforts by the worldwide research community to challenge and improve the existing modelling platform.

With the CRI in operation for over two years, the public good approach to credit ratings is no longer just a concept. The availability of free and transparent credit information to all could lead to the much needed overhaul in the credit ratings industry. The authors conclude that the public good alternative is arguably the most promising credit rating reform idea that has been contemplated thus far and that the CRI’s potential to reshape the credit rating industry is real.

Back To Newsletter

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