Bond-CDS Implied Rating Systems
34 Pages Posted: 25 Nov 2014 Last revised: 24 Mar 2020
Date Written: October 20, 2019
Abstract
Market-implied ratings gained importance as efficient early warnings of official credit rating migrations. We build a two-dimensional implied rating system that gathers information from both the bond and the CDS markets. The system is able to outdo each of the corresponding one-dimensional implied ratings in terms of forecast performances. This allows to bypass the debate on which of the two markets can better anticipate eventual rating migrations, since the information is efficiently conveyed in a unique implied rating system. We show that this simple method is also able to outperform implied rating systems based on recent machine learning techniques.
Keywords: credit ratings, market implied ratings, support vector machine, machine learning, credit default swaps, default risk
JEL Classification: C38, D83, G17, G24
Suggested Citation: Suggested Citation