Inteligência computacional aplicada na análise e recuperação de portfólios de créditos do tipo non-performing loans

• Ano de defesa: 2015
• Autor(a) principal: Souza, Flávio Clésio Silva de
• Orientador(a): Sassi, Renato José
• Banca de defesa: Dias, Cleber Gustavo , Souza, Reinaldo Castro
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Nove de Julho
• Programa de Pós-Graduação: Programa de Pós-Graduação de Mestrado e Doutorado em Engenharia de Produção
• Departamento: Engenharia
• País: Brasil
• Palavras-chave em Português: non-performing loans; fundos de investimento em direitos creditórios; inteligência computacional; tomada de decisão; créditos não-perfomados
• Palavras-chave em Inglês: non-performing loans; fundos de investimento em direitos creditórios; computational intelligence; decision making; default loan
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA DE PRODUCAO
• Link de acesso: https://bibliotecatede.uninove.br/handle/tede/1012

One of the economic externalities for the credit increase is the resulting increase in the default rate. In the face of this economic scenario emerged in Brazil financial and banking institutions are offer through the sale of such claims. These defaulted credits are called Non-Performing Loans (NPLs). The demand to the purchase of NPLs is made by economic structures called Fundos de Investimentos em Direitos Creditórios (FIDC), that are aimed to get financial return through the recovery of such credits. The funds perform several analysis of business viability through statistical techniques, financial, and economic conditions in search of the determinants that influences directly in the recovery of these credits. However, other techniques of the computational intelligence can be applied in the analysis of such credits. This main objective of this work is the application of Computational Intelligence techniques in the recovery of Non-Performing Loans. Studies in the literature, yet not directly address questions of how these analyzes directly influence the evaluation capacity or pricing of these financial assets, the recovery of credits based on the already defaulted credit perspective; and also what are the determinants that influence the recovery of these credits. Studies in the literature, yet not directly address questions of how these analyzes directly influence the evaluation capacity or pricing of these financial assets, the recovery of credits based on the already defaulted credit perspective; and also what are the determinants that influence the recovery of these credits. Three experiments were conducted using the following Computational Intelligence techniques: Artificial Neural Networks, Theory of Rough Sets and Decision Trees. The results obtained with the application of techniques were conclusive to point out that the factors related to the forms of location of debtors, aging, and the value of debt are the main determinants in the recovery of NPLs; and therefore must be taken into account in decision support is for evaluation activities and pricing of these assets, is to prepare recovery strategies of these assets.