The Use of Accounting Data to Assess Default Risk in Credit Operations of Small Companies

Abstract

The purpose of this article is to verify the possibility of evaluating the credit risk of small businesses based on accounting data for a sample of solvent and insolvent loans, by developing an early warning model. Based on this, the paper sought to evaluate the results obtained from each of the selected accounting indicators according to the criterion of relevance established by the author, also verifying if the result was expected or not. The sample is created with data from 2006, which are used for the analysis. This fact should be emphasized since insolvency forecast models lose their validity over time due to changes in the economic scenario. The model was created to evaluate the risk. One innovation of this article is the criteria to split the sample in two groups, which are created, based on criteria established by the National Monetary Council. The objective of these analyses is to anticipate the probability of credit risk migration, especially in the case of deterioration in the risk of bankruptcy. Key-worlds: Credit risk; discriminant analysis; bankruptcy; early warning.