Can information alleviate overconfidence? A randomized experiment on financial market predictions
Abstract: In this study, we investigate how providing information affects the level of overconfidence among participants through an online experiment. Participants in stock market prediction exercises were asked to assess their absolute and relative performance. We conducted a randomized controlled trial in which selected participants received information about their own performance and/or the performance distribution of others. The results show that, although participants tend to be overconfident on average, their responses exhibit substantial heterogeneity. Information provision reduces the bias on average. The reduction in overconfidence is more pronounced when participants are given partial rather than complete information. The results also suggest that simply providing information-regardless of whether it aligns with participants' prior beliefs-decreases confidence about their performances even when initially underconfident. The belief regarding the performance distribution of others is not affected by information provision.
Register in advance for this meeting: here.
After registering, you will receive a confirmation email containing information about joining the meeting.
The virtual seminar will consist of a 60-minute research paper presentation, followed by 15 minutes of Q&A and 15 minutes of discussion with students/junior researchers. During the final 15-minute discussion session, we encourage students and junior researchers to stay and interact with the speaker.