Re-Examining Young Adults' Financial Overconfidence

Abstract

This study develops an alternative method for identifying financial overconfidence among young adults. Using data from the 2021 National Financial Capability Study (NFCS), we apply latent class analysis (LCA) to classify individuals based on multiple dimensions of financial knowledge: self-assessed knowledge, objective responses, and “Don't Know/Refuse” (DK/RF) selections. Unlike approaches based on discrepancy scores or median splits, LCA models the joint distribution of these indicators and assigns individuals to latent profiles, allowing heterogeneity in knowledge and response behavior to be retained. The model identifies four profiles: Consistently High Knowledge, Overconfident, Reluctant, and Consistently Low Knowledge. The Overconfident class reflects high subjective knowledge with lower objective accuracy, while the Reluctant class is characterized by frequent DK/RF responses. The Overconfident group shows the highest likelihood of alternative financial services (AFS) usage, whereas the Consistently High Knowledge group shows the lowest. These findings indicate that classification strategy shapes how overconfidence is defined and related to financial behavior.