Current literature provides various models for credit risk analysis, called Credit Scoring Models (CSMs). However, these models are not suited to the majority of Micro and Small Enterprises (MSEs). This is compounded by a lack of technical knowledge of microentrepreneurs linked to the high costs and complexity of the CSMs. These issues are significant as 99% of Brazilian companies are MSEs. Therefore, this paper aims to propose a CSM for an MSE that commercializes construction materials in São Paulo, Brazil. This research is quantitative and characterized as a case study whose CSM is based on the Naive Bayes algorithm implemented in Microsoft Office Excel 2016. This model calculates the probability of default and adherence by weighting the results based on the Modern Finance Theory with the Cost of Denying and the Cost of Granting. The application of the model demonstrates that the successes in approvals (70%) and disapprovals (66%) of Credit Sales were significant with a result of R$ 32.20 thousand and an increase in net profit by 124.2%. We have evidenced that the proposed CSM is able to weigh the risk in investment and financing decisions for the Credit Sales of an MSE. This paper provides a low-cost CSM, adapted to reality and easy to handle and implement in MSEs. This research is a reference to the development of CSMs focused on the credit concessions conducted by MSEs.
Read full abstract