Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Market-Based Evaluation for Models to Predict Bond Ratings

Market-Based Evaluation for Models to Predict Bond Ratings Previous studies have examined different statistical models to predictcorporate bond ratings. However, these papers use agency ratings as thebenchmark to assess models and ignore the evidence that agency ratings maynot be accurate in a timely manner. In this paper, we propose a new approachwhich incorporates ex-post bond returns to evaluate rating prediction models.Relative rating strength portfolios, formed by buying under-rated bonds withagency ratings lower than model ratings and selling over-rated bonds withagency ratings higher than model ratings, are employed to test the performanceof different statistical models in rating predictions. Our results show thatone version of multiple discriminant analysis model can generate a statisticallysignificant abnormal return of 5% over a 5-year horizon. The ordered probit modelwhich is believed to possess theoretical advantages in classifying bonds doesnot perform better. This suggests that using traditional measures to evaluatemodels can be misleading. The existence of a profitable trading strategy alsoraises the concern of market efficiency in the corporate bond market. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Review of Pacific Basin Financial Markets and Policies World Scientific Publishing Company

Market-Based Evaluation for Models to Predict Bond Ratings

Loading next page...
 
/lp/world-scientific-publishing-company/market-based-evaluation-for-models-to-predict-bond-ratings-VwQ9MI9Cw8

References (37)

Publisher
World Scientific Publishing Company
Copyright
Copyright ©
ISSN
0219-0915
eISSN
1793-6705
DOI
10.1142/S0219091504000081
Publisher site
See Article on Publisher Site

Abstract

Previous studies have examined different statistical models to predictcorporate bond ratings. However, these papers use agency ratings as thebenchmark to assess models and ignore the evidence that agency ratings maynot be accurate in a timely manner. In this paper, we propose a new approachwhich incorporates ex-post bond returns to evaluate rating prediction models.Relative rating strength portfolios, formed by buying under-rated bonds withagency ratings lower than model ratings and selling over-rated bonds withagency ratings higher than model ratings, are employed to test the performanceof different statistical models in rating predictions. Our results show thatone version of multiple discriminant analysis model can generate a statisticallysignificant abnormal return of 5% over a 5-year horizon. The ordered probit modelwhich is believed to possess theoretical advantages in classifying bonds doesnot perform better. This suggests that using traditional measures to evaluatemodels can be misleading. The existence of a profitable trading strategy alsoraises the concern of market efficiency in the corporate bond market.

Journal

Review of Pacific Basin Financial Markets and PoliciesWorld Scientific Publishing Company

Published: Jun 1, 2004

Keywords: Bond rating prediction relative rating strength portfolio bond trading strategy bond market efficiency

There are no references for this article.