Abstract
This study examines how overconfidence—a cognitive bias proxied by the joint condition of labor informality and credit use—affects the probability of household over-indebtedness in Metropolitan Lima during 2019–2023, controlling for temporal myopia and sociodemographic factors. Using harmonized ENAHO household–year microdata (20,332 observations), over-indebtedness is defined as a binary outcome when households report credit use and a debt-service-to-income ratio above 0.30, yielding an incidence of 0.98% and characterizing the event as rare. The results show that overconfidence is strongly and statistically associated with over-indebtedness (odds ratio above 70; p<0.001), increasing predicted probability from approximately 0.5% to nearly 29%, holding controls constant. Additional evidence reveals a non-linear life-cycle pattern in age and a significant temporal effect in 2023, consistent with elevated financial stress. In a robustness model restricted to credit users, overconfidence loses significance while per capita income emerges as a protective factor, indicating that behavioral bias primarily operates as an ex-ante vulnerability profile linked to informality and credit exposure, whereas repayment capacity dominates once credit access is conditioned. These findings underscore the relevance of behavioral economics in explaining rare but high-risk financial outcomes and support the incorporation of behavioral early-warning indicators into credit risk management and preventive policy design.

This publication has 21 references indexed in Scilit: