Uso da análise de sobrevivência na avaliação de ativos contratuais

Abstract

The application of survival analysis in the evaluation of contractual assets within the context of banking correspondents (CORBAN) was investigated, highlighting the relevance of this technique to model and forecast future financial flows, ensuring the economic sustainability of these entities. Using a comprehensive database of payroll loan contracts and the RStudio software, parametric models such as Weibull and log-normal, as well as semi-parametric models like the Cox proportional hazards model, were explored to analyze contract duration and identify determinant variables. The results showed that most failures occur within the first 24 months, with Weibull models presenting a good fit to describe the survival function, while the Cox model confirmed the significant influence of covariates such as financing bank, contract type, loan line, and group. It was observed that contract survival varies according to these characteristics, providing important insights for financial planning and pricing strategies. The research also underscores the need for statistical tools to support the economic and financial management of contractual assets. Survival analysis proved effective in providing quantitative bases for decision-making and suggests future refinements to incorporate dynamic changes in contracts and their variables.