Construcción de un modelo scoring de aprobación para cartera comercial de una institución financiera pública mediante modelos aditivos generalizados

Abstract

This project aims to analyze the application of a generalized additive model in the financial field, given the need of financial institutions to improve the placement of their credit portfolio in good clients, that is, that they can meet their obligations agreed with the institution in the foreseen terms, without incurring partial or total delays, thus allowing to minimize their credit risk. In this sense, the theoretical framework has been incorporated into this type of statistical models, credit risk, an analysis of the current situation of Ecuadorian banking, to later propose a methodological procedure that allows applying a generalized additive model for the portfolio of commercial loans in a public development bank, in order to establish strategies for credit granting. Programming was performed in software R. This project proposes the creation of new credit scoring models whose functional structures contemplate the possible non-linearity of the explanatory variables of credit risk in relation to compliance with the payment obligations of the borrowers.