In this study, we examine the predictive performance of a wide class of binary classifiers using a large sample of international credit ratings changes from the period 1983–2013. Using a number of financial, market, corporate governance, macro-economic and other indicators as explanatory variables, we compare classifiers ranging from conventional techniques (such as logit/probit and LDA) to fully nonlinear classifiers, including neural networks, support vector machines and more recent statistical learning techniques such as generalised boosting, AdaBoost and random forests. We find that the newer classifiers significantly outperform all other classifiers on both the cross sectional and longitudinal test samples; and prove remarkably robust to different data structures and assumptions. Simple linear classifiers such as logit/probit and LDA are found nonetheless to predict quite accurately on the test samples, in some cases performing comparably well to more flexible model structures. We conclude that simpler classifiers can be viable alternatives to more sophisticated approaches, particularly if interpretability is an important objective of the modelling exercise. We also suggest effective ways to enhance the predictive performance of many of the binary classifiers examined in this study.

/pdf/an-empirical-evaluation-of-the-performance-of-binary-2isx1qi7l3.pdf

An empirical evaluation of the performance of binary classifiers in the prediction of credit ratings changes

Diversification and bank stability in the GCC

This paper examines the capability of firm fundamentals in explaining the cross-sectional variation of credit default swap (CDS) spreads. The paper constructs a fundamental-based CDS valuation by combining the Merton distance-to-default measure with a long list of firm fundamental characteristics. Regressing market CDS quotes against the fundamental valuation cross-sectionally generates an average R-squared of 77%. The cross-sectional explanatory power is stable over time, and robust in out-of-sample tests. Deviations between market quotes and the fundamental valuation predict significantly future market movements. The results highlight the important role of firm fundamentals in differentiating the credit quality of different firms.

Anchoring credit default swap spreads to firm fundamentals

This paper uses a sample of 2,186 credit default swap spreads quoted in the European market during the period 2002–2009 to empirically analyze which model – accounting- or market-based – better explains corporate credit risk. We find little difference in the explanatory power of these two approaches. Our results indicate that a comprehensive model that combines accounting- and market-based variables is the best option to explain the credit risk, suggesting that both types of data are complementary. We also demonstrate that the explanatory power of credit risk models is particularly strong during periods of high uncertainty, such as those experienced in the recent financial crisis. Finally, the comprehensive model continues to produce the best results if the credit rating is used as the proxy for credit risk; however, accounting variables currently appear to have a more important role than market variables in determining corporate credit ratings.
First published online:  05 Feb 2013

/pdf/examining-what-best-explains-corporate-credit-risk-39uxupedw4.pdf

Examining what best explains corporate credit risk: accounting-based versus market-based models

Ratings issued by credit rating agencies (CRAs) play an important role in the global financial environment. Among other issues, past studies have explored the potential for predicting these ratings using a variety of explanatory factors and modeling approaches. This paper describes a multi-criteria classification approach that combines accounting data with a structural default prediction model in order to obtain improved predictions and test the incremental information that a structural model provides in this context. Empirical results are presented for a panel data set of European listed firms during the period 2002–2012. The analysis indicates that a distance-to-default measure obtained from a structural model adds significant information compared to popular financial ratios. Nevertheless, its power is considerably weakened when market capitalization is also considered. The robustness of the results is examined over time and under different rating category specifications.

/pdf/combining-accounting-data-and-a-structural-model-for-1df5mfrjy4.pdf

Combining accounting data and a structural model for predicting credit ratings: Empirical evidence from European listed firms

We investigate the empirical performance of default probability prediction based on Merton's (1974) structural credit risk model. More specifically, we study if distance-to-default is a sufficient statistic for the equity market information concerning the credit quality of the debt-issuing firm. We show that a simple reduced form model outperforms the Merton (1974) model for both in-sample fitting and out-of-sample predictability for credit ratings, and that both can be greatly improved by including the firm's equity value as an additional variable. Moreover, the empirical performance of this hybrid model is very similar to that of the simple reduced form model. As a result, we conclude that distant-to-default alone does not adequately capture the firm's credit quality information from the equity market. Copyright © 2007 ASAC. Published by John Wiley & Sons, Ltd.



Dans cet article, nous evaluons la performance empirique de la prediction de probabilite de non-remboursement a partir du modele de risque de credit structurel de Merton (1974). Plus particulierement, nous cherchons a examiner dans quelle mesure la “distance par rapport a la defaillance” (distance-to-default) est une statistique suffisante pour l'information du marche d'actions concernant la qualite du credit de la compagnie emettrice. Nous demontrons qu'un modele simple de formule reduite donne de meilleurs resultats a la fois pour l'agencement echantillonne et pour la predictibilite hors-echantillon que le modele de Merton (1974), et que ces deux resultats peuvent etre encore plus precis si la valeur comptable de la firme est prise en compte comme variable supplementaire. Par ailleurs, la performance empirique de ce modele hybride est tres semblable a celle du modele simple de forme reduite. Aussi concluons-nous que la distance par rapport a la defaillance ne rend pas fidelement compte de l'information concernant la qualite du credit provenant du marche d'actions. Copyright © 2007 ASAC. Published by John Wiley & Sons, Ltd.

Assessing credit quality from the equity market: can a structural approach forecast credit ratings?

In this paper, we investigate whether the theoretical default probability measures calculated from Merton's (1974) structural credit risk model can provide a better way to explain and predict credit rating than traditional statistical models. The empirical results suggest that Merton's theoretical default measure is not a sufficient statistic of equity market information concerning credit quality. By including the market value of the firm as an independent variable we can improve both in-sample fitting and out-of-sample predictability of credit ratings. Moreover, the empirical performance of this hybrid model is very similar to those simple statistical model. As a result, we conclude that structural models hardly provide any additional capability in capturing credit risk. Our empirical results show that instead of using the firm value only through the debt leverage ratio, as suggested in the structure models, one should include the market value of the firm as a separate factor affecting default probability when building credit risk models.

https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID470701_code97747.pdf?abstractid=470701&mirid=3

Assessing Credit Quality from Equity Markets: Is Structural Model a Better Approach?

Taking advantage of the fexibility of the simple hazard rate model estimation technique similar to Shumway (2001), this paper studies the effects of duration, momentum, and rating policies on credit rating changes. In contrast to previous findings, we find that the duration effect on credit rating changes is not monotonic. Further, we document that the duration effect on downgrade is mainly caused by the downgrade momentum.

Yu Du

Papers

Assessing credit quality from the equity market: can a structural approach forecast credit ratings?

Assessing Credit Quality from Equity Markets: Is Structural Model a Better Approach?

An Empirical Study on Credit Rating Change Behavior