Credit ratings play an important role in financial markets: regulators use credit rating to monitor the financial health of banks, pension funds, and other financial institutions; investors use credit rating to assess the riskiness of their investment and adjust their portfolios accordingly; banks employ credit rating migration metrics to calculate the default correlation among the assets in their portfolios. Moreover, derivatives whose payoffs are directly linked to the credit ratings of some reference firms have become increasingly popular.
Credit ratings are ordinal measures to reflect firms’ credit quality. The rating process includes quantitative as well as qualitative and legal analysis. While similar approaches are used by different rating agencies when assigning credit ratings, and some agencies are more forthcoming than others in describing the procedures they follow in assigning or reviewing a rating, they all use proprietary methods to do so. For these reasons, various models have been built to predict a firm’s credit rating in the future based on publicly available information. Researches in this area include Horrigan (1966), West (1973), Pogue and Soldofsky (1969), Kaplan and Urwitz (1979), and Blume, Lim and Mackinlay (1998), Nickell et al (2000), among others. Despite the skepticism from rating agencies, these models have been fairly successful in explaining and predicting the ratings of a large cross section of corporate bonds.
Most of the studies concerning credit rating forecasting have focused on either determining the factors affecting a firm’s credit rating, or building better econometric techniques. Popular approaches include hazard rate models, logistic type of models, and discriminant analysis, where usually the chosen independent variables enter the estimation process in a linear way. Specifically, little attention has been paid to the possible structural interactions among the independent variables, and to the possible nonlinear effect of the covariates. With the development of option pricing theory, researchers have been able to build structural models to price corporate debt by treating it as a derivative written on the firm’s underlying assets. Since the seminal work of Merton (1974), many such models have been developed. Using equity market data, they can generate a quantity representing the default probability of each particular firm. Unlike those traditional reduced form statistical models, structural models relate different credit risk factors in an analytical way, and it is natural to ask whether the default probability inferred from structural credit risk models can provide a better alternative to assessing firms’ credit quality.
Empirical studies on structural credit risk models generally focus on the relationship between the default risk with corporate bonds yields (see Eom et al (2004) and the reference therein). Few empirical studies have been conducted on the relationship between actual default frequency and theoretical default probability calculated from these models. Notable exceptions include Hillegeist et al (2004), and Vassalou and Xing (2004). Hillegeist et al (2004) find that structural default probability measures contain relatively more information than Altman’s Z-Score and Ohlson’s O-Score. Using measures calculated from structural credit risk models as proxy to the firms’ default risk, Vassalou and Xing (2004) find that the Fama-French factors SMB and HML (see Fama and French (1996)) do not proxy the default factors. Moreover, in practice Moody’s KMV has been using the Merton type of structural models to predict the default probabilities of individual firms.
Download
Assessing Credit Quality from Equity Markets: Can Structural Approach Forecast Credit Ratings?
