A MODEL FOR THE GRANTING OF CREDITS AND RISK ESTIMATION IN THE AGRICULTURAL SECTOR

Abstract

This study aims to propose a statistical model for the granting of credits in the agricultural sector that will generate an appropriate credit risk management. (Case of the Trust Funds to Agriculture, FIRA). The methodology used was the discrimination technique, commonly known as credit scoring.The score tables allow to determine default probabilities of guaranteed loans, these probabilities and the contingent balance are inputs to estimate credit risk through three models commonly used in practice: CyRCE, Montecarlo and Credit Risk+. The results led to the conclusion that the parameters determined through a scoreboard based on logistic regression and estimation of credit risk with the Monte Carlo model, allow having a balance between revenue and expenditure.