SURVIVAL ANALYSIS: DEFAULT RISK IN A CREDIT UNION IN RONDÔNIA

The objective of this work was to apply survival analysis to investigate and model the time until default occurs among clients of a credit cooperative in Rondônia. It describes the temporal behavior of default rates and identifies factors associated with credit risk, using survival estimates. It details the Kaplan-Meier methodology, explaining its application to estimating the survival function and comparing survival curves between different groups of clients, categorized by socioeconomic and financial variables, allowing the identification of patterns and critical periods of higher risk of default. Statistical tests, such as Log-rank, are used to assess the significance of differences between survival curves. The results demonstrate that certain customer characteristics and loan conditions are strongly associated with the risk of default. Based on Kaplan-Meier probability estimates, the study provides insights into the periods when customers are most likely to default. In conclusion, the application of survival analysis proved to be an efficient tool for demonstrating which variables have a high risk of leading to default, contributing to the improvement of credit management practices in financial cooperatives by providing quantitative evidence on risk factors for default and for mitigating these risks.