Classificação de risco de crédito: modelos estruturais, modelos não estruturais, ratings das agências de classificação: convergências e/ou divergências?

Abstract

The corporate credit risk is the risk which the lending institution is exposed if some of its counterparties fail in fulfilling the contractual obligations for credit. For academics, a couple of methodologies have been used to estimate the credit risk of companies: the structural and the non structural models. This research focuses the KMV structural model and the non structural models from Kanitz (1976), Altman, Baidya and Dias (1979), Minussi (2008) and Brito & Assaf Neto (2008), in order to verifying the level of convergence between the results estimated through these methodologies and compare the classification obtained for them with the rating given by the Moody?s and Standart & Poor?s rating agencies. The study was developed through an explanatory research within the period of 2006 to 2009. Simple and multiples linear regressions were estimated in order to verify the convergence between the models and a comparative analysis between the classifications obtained through the models and the ratings from those agencies. It was concluded that the results estimated through non structural model of Altman, Baidya and Dias (1979) are convergent with the results obtained through the KMV structural model. The results estimated through the Brito and Assaf Neto (2008) model were convergent with the KMV model when analyzed together with the results from another models. The results obtained through the Kanitz (1976) and Minussi (2008) models did not show convergence. Regarding to the comparisons between the results obtained for the models and the ratings of agencies, the KMV structural model and the non structural models of Minussi (2008) and Altman, Baidya and Dias (1979) stood out for having the most similar results.