LOGISTIC REGRESSION AND GENETIC ALGORITHMS APPLIED TO CREDIT RISK ANALYSIS
The taking of decisions of credit concession is based basically on the evaluation of the insolvency risk of potential contractors of credit products. With the technological advance, statistical models have been developed to support the analysis of credit requests, which was many times carried throug...
| Autores: | , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2012 |
| País: | Brasil |
| Institución: | Universidade Regional de Blumenau (FURB) |
| Repositorio: | Revista Universo Contábil |
| Idioma: | portugués |
| OAI Identifier: | oai:ojs.bu.furb.br:article/2374 |
| Acceso en línea: | https://ojsrevista.furb.br/ojs/index.php/universocontabil/article/view/2374 |
| Access Level: | acceso abierto |
| Palabra clave: | Credit risk. Credit scoring models. Logistic regresion. Genetic algorithms. Riesgo de crédito. Modelos de credit scoring. Regresión logística. Algoritmos genéticos. Risco de crédito. Modelos de credit scoring. Regressão logística. Algoritmos genéticos. |
| Sumario: | The taking of decisions of credit concession is based basically on the evaluation of the insolvency risk of potential contractors of credit products. With the technological advance, statistical models have been developed to support the analysis of credit requests, which was many times carried through qualitatively some decades ago. The goal of this study is to present the use of logistic regression and genetic algorithms for sorting good and bad payers in bank financing and the identification of the best model in terms of goodness-of-fit. From a sample of 14,000 data, supplied by a great Brazilian financial institution, the two techniques were applied. Logistic regression presented the best goodness-of-fit. This work illustrated the procedures to be adopted by a company to identify the best model of credit concession, from which it is possible to direct the strategy of the institution in the evaluation process of bank loan requests. |
|---|