This paper investigates the causal effect of textual framing on risk classification in Investor Profile Analysis (IPA) questionnaires in Brazil. We tested whether the formulation of the questions systematically alters the classification of the investor's risk profile by conducting a randomized controlled trial (RCT) with 900 Brazilian adults, allocated to one of three versions of the questionnaire: Neutral (control), Positive (framing in gains), and Negative (framing in losses). The results reveal that negative framing increases the probability of classification as Conservative by 11 percentage points, while positive framing increases the classification as high-risk by 9 percentage points. The magnitude of the effect represents a change of approximately half a standard deviation in risk scores. Our results suggest that the presentation format of the suitability questionnaires can affect not only risk perception but also the consistency of investors' self-reported preferences.
Comissão Organizadora
Comissão Científica