Two Measures of Financial Risk Tolerance from Questionnaire Data

Abstract

Set-based principal component analysis extracts two independent measures of financial risk tolerance from questionnaire responses. These measures are less noisy than most scoring methods. They are also more robust because they provide a redundant assessment of financial risk tolerance. This analysis reveals that typical risk questionnaires measure financial risk tolerance and income-related risk tolerance. These two factors are unrelated, which suggests that income-related questions should be dropped from financial risk questionnaires. We also examine how financial risk tolerance depends on financial literacy and gender.