| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.14 MB | Adobe PDF |
Orientador(es)
Resumo(s)
Com intenção de melhorar os serviços de análise e gestão de crédito, as instituições financeiras desenvolveram o modelo credit scoring. Este modelo é utilizado por estas instituições para previsão do risco de crédito no processo da tomada de decisão de concessão de crédito. O objetivo deste trabalho, é desenvolver um modelo de credit scoring a partir de uma amostra de 1000 solicitantes de créditos extraídos da carteira de crédito de um banco alemão. Para tal, estimou-se um modelo probit, considerando-se 25 variáveis independentes quantitativas e qualitativas que influenciam a probabilidade do crédito ser aprovado ou não. Os resultados deste estudo mostram que o modelo de credit scoring se apresenta adequado no ajustamento aos dados, obtendo uma classificação correta para cerca de 77% dos clientes. Contudo, os resultados encontrados fornecem informações importantes para auxílio no processo de tomada de decisões de concessão de crédito e gerenciamento do crédito bancário, podendo assim contribuir para a redução do número de clientes inadimplentes e dos respetivos custos.
In order to improve credit analysis and management services, financial institutions have developed the credit scoring model. This model is used by these institutions to predict credit risk in the process of making a credit granting decision. The objective of this work is to develop a credit scoring model from a sample of 1000 credit claimants extracted from the credit portfolio of a German bank. For this, a probit model was estimated, considering 25 independent quantitative and qualitative variables that influence the probability of credit being approved or not. The results of this study show that the credit scoring model is adequate in the adjustment to the data, obtaining a correct classification for about 77% of the clients. However, the results found provide important information to aid in the decision-making process of credit granting and bank credit management, thus contributing to the reduction of overdue customers and their costs.
In order to improve credit analysis and management services, financial institutions have developed the credit scoring model. This model is used by these institutions to predict credit risk in the process of making a credit granting decision. The objective of this work is to develop a credit scoring model from a sample of 1000 credit claimants extracted from the credit portfolio of a German bank. For this, a probit model was estimated, considering 25 independent quantitative and qualitative variables that influence the probability of credit being approved or not. The results of this study show that the credit scoring model is adequate in the adjustment to the data, obtaining a correct classification for about 77% of the clients. However, the results found provide important information to aid in the decision-making process of credit granting and bank credit management, thus contributing to the reduction of overdue customers and their costs.
Descrição
Mestrado em Métodos Quantitativos para a Decisão Económica e Empresarial
Palavras-chave
Crédito Risco de Crédito Credit Scoring Probit Credit Credit risk
Contexto Educativo
Citação
Fernandes, António Francisco de Melo (2017). "Credit scoring : uma análise econométrica". Dissertação de Mestrado, Universidade de Lisboa. Instituto Superior de Economia e Gestão.
Editora
Instituto Superior de Economia e Gestão
