Cartão de crédito: prevendo a taxa de inadimplência através dos indicadores macroeconômicos

• Ano de defesa: 2025
• Autor(a) principal: Murtadha, Nabil Ahda
• Orientador(a): Cândido, Osvaldo
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Católica de Brasília
• Programa de Pós-Graduação: Programa Stricto Sensu em Economia de Empresas
• Departamento: Escola de Humanidades, Negócios e Direito
• País: Brasil
• Palavras-chave em Português: Inadimplência; Cartão de crédito; Variáveis macroeconômicas; Modelo ARDL; Aprendizado de máquina; Risco de crédito
• Palavras-chave em Inglês: Credit default; Credit cards; Macroeconomic variables; ARDL model; Cointegration; Machine learning; Credit risk
• Área do conhecimento CNPq: CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA
• Link de acesso: https://bdtd.ucb.br:8443/jspui/handle/tede/3702

This study investigates the macroeconomic determinants of credit card default in Brazil, based on monthly data from March 2011 to May 2024. The ARDL (Autoregressive Distributed Lag) methodology is used to identify short- and long-run relationships between default rates and macroeconomic variables. The results indicate that credit card default is influenced by macroeconomic factors, particularly the Selic interest rate, inflation, exchange rate, and credit volume. GDP showed a significant effect only in the short run. Differences between short- and long-run effects were observed, especially regarding inflation and credit, suggesting a complex dynamic. Machine learning models were applied to complement the analysis, confirming the relevance of the selected variables and demonstrating strong predictive performance. The convergence between econometric and predictive approaches reinforces the robustness of the results and their usefulness for monitoring credit risk in the credit card segment.