Un modelo de predicción para analizar el riesgo de crédito en MiBanco S.A.

Abstract

In Colombia, the majority of commerce is sustained by microenterprises, which play a vital role in production and job creation. To ensure their optimal functioning, financial institutions have developed operational models based on offering financing and savings solutions through microcredits. Financial inclusion requires a precise analysis of credit risk and the calculation of indicators to determine the most suitable credit solution for both the microenterprise and the financial institution. This work proposes the use of microenterprise databases, for data security reasons, the models and results presented are based on simulated data. A characterization and prediction analysis of credit access is conducted, selecting significant variables and making curve adjustments using the least squares method, including regression lines, potential fitting, and data linearization, with the aim of finding optimal curves that adapt to data behavior. The conclusion presents the optimal models and a comparative analysis. To evaluate the models and validate the techniques used, the practice was carried out at MiBanco S.A.