If you are unable to see the message below, click here to view.  
HOME Recent Events RMI in the News ALUMNI
  Issue 7 | Archive  

May 2011

RMI Symposium Discusses Implementation of Credit Rating Systems

With the latest industry developments such as the Dodd-Frank Act in the U.S., which requires regulators to remove references to credit rating agency ratings from their regulations, implementing robust and informative credit rating systems has become increasingly important. In response, on 29 March 2011, RMI held a half-day Symposium on Credit Rating Systems to confront the implementation issues, drawing a crowd of more than 100 industry practitioners and academics.

Starting the symposium off was Mr. Martin Fassbender, Head of the Eligible Assets Section and the Credit Risk Analysis Department from Deutsche Bundesbank (BBk), German's central bank. Together with his colleague, Dr.Laura Auria, Deputy Head of the Credit Risk Analysis Department, they presentedto the audience BBk's Credit Risk Analysis as Part of the Eurosystem Credit Assessment Framework (ECAF), which was set up to guarantee a consistent framework for the evaluation of the credit quality and ensure that high credit standards for all eligible assets are met. Their presentation gave an overview on the elements of the ECAF framework and laid down the environment, in which the BBk credit risk analysis system works. They also shared the main elements of BBk's risk analysis process.

Second speaker in the spotlight was Mr. Benjamin Wong, Head of Risk Analytics from Standard Chartered Bank, Singapore. He introduced a macroeconomic factor based Point-in-Time (PiT) rating system. He pointed out that neither pure PiT nor Through-the-Cycle (TtC) ratings systems exist in practice. Instead, most actual rating systems are "Hybrid" and predict somewhere between PiT Probability of Default (PD) and TtC PD depending on the correlations between model factors and macro-economic conditions. He suggested that by quantifying the systematic risk and model cyclicality of a rating system,a macroeconomic model can be built to forecast PiT PD. This can also be extended to macroeconomic scenario based stress testing, he added.

His talk was followed by BBk's Dr. Auria's presentation on support vector machines (SVM) for credit risk analysis. She gave an overview on the main elements of the SVM technique. After showing the role of the kernel for making data, which are non-linearly separable in an input space, linearly separable in a feature space, she also explained how the techniques solved the problem of model risk due to non-representative data with the help of generalization parameters.

Subsequently, Dr. Ruslan Moro, lecturer from the Brunel University in the U.K., and Senior Economist at DIW econ, the consulting company of Germany's DIW Berlin, showed to the audience the analysis of corporate default in 12 countries of the Asia-Pacific region based on parametric (logit) and non-parametric (SVM) techniques. He also briefly compared the characteristics of insolvent companies in Asia and Europe (Germany) before ending his session with a discussion on the connection between innovation activity and credit rating where innovation activity were measured based on the number of patent applications that a company filed worldwide.

In the symposium's final session, Dr. Oliver Chen, RMI's Senior Research Fellow presented on the performance analysis of RMI Credit Rating Initiative (CRI)'s PD forecasts. He introduced the CRI's current model and addressed some of the specific implementation details, especially for assessing the financial sector and smaller economies. Detailed economy-by-economy and sector-by-sector performance analysis were also presented for economies under CRI's current coverage, including the 12 Asia-Pacific economies as well as the U.S. and Canada. Possible refinements and enhancements to the model were also discussed.

Back To Newsletter

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