Análise de curto e longo prazo? uma alternativa às metodologias de rating de crédito convencionais no Brasil

Abstract

Credit Rating Agencies (CRA) apply in their rating assignments a methodology that filters out the effect of short-term fluctuations in credit risk, while alternative methodologies based on quantitative approaches do not suppress this short-term component that are potentially relevant in secondary market transactions. Using a variable of short-term fluctuations as a proxy for credit risk, we observe that ratings assigned by one of those alternative methodologies change more often than ratings assigned by CRAs, and we also found statistically significant evidences of the relationship between those short term fluctuations and credit spreads on the secondary market for brazilian companies, this way filling an existing research gap in this topic for the domestic market. Aditionally, when the short term proxy was tested together with ratings assigned by CRAs, we observed improvements on the model explanatory power. Our data sample period ranges from 2015 to 2019, with a total of 180 securities. Data was organized in panel and tests were performed using fixed and random effects, with superior goodness-of-fit observed in the first approach.