Captura de expertise de analistas em processo de pré-análise de crédito

• Ano de defesa: 2020
• Autor(a) principal: Ribas, Ariane Mayra de Castro Campos Taborda
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Curitiba; Brasil; Programa de Pós-Graduação em Administração; UTFPR
• Palavras-chave em Português: Análise de crédito; Pequenas e médias empresas; Sistemas de avaliação de risco de crédito (Finanças); Cooperativas de crédito; Microfinanças; Simulação (Computadores); Créditos - Avaliação; Credit analysis; Small business; Credit scoring systems; Banks and banking, Cooperative; Microfinance; Computer simulation; Credit ratings; CNPQ::CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO; Administração
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/25356

Guarantee Societies are organizations whose objective is to mitigate imperfections in the microcredit market in terms of supplying resources to Micro and Small Enterprises through the provision of guarantees. The main difficulties faced by MSEs are high interest rates, information asymmetry and the absence of collateral. Credit Guarantee Societies are different from other financial institutions in that they have both economic and social objectives. Where they seek to make the return for members viable by offering financial services and more favorable conditions. In addition to reaching more distant and less economically favorable regions. The present study aims to develop and adjust a credit scoring model for pre-analysis of credit guarantee operations based on the expertise of analysts from Curitiba Credit Guarantee Societies. For this purpose, the logistic regression method was used, estimating the parameter by maximum likelihood, data validation was done by K-fold Cross Validation and legitimation by ACC, MCC, Precision, Recall and F1-score. Being performed in software R. The estimated model had statistical measures of performance greater than 75% for four of the five metrics analyzed. The results enable SGCs’ credit analysts to automate the analysis based on their own expertise, preparing for the increased demand for microcredit, as has been happening in recent years.