Overconfidence, position size, and the link to performance

The overconfidence literature employs activity metrics such as account turnover and trade frequency to link misattribution/self-attribution to excess trading. In this paper we argue relative position size is a more meaningful indicator of overconfidence. Using a sample of retail traders, we find that when traders take relatively larger positions they make more impaired trade entry/exit timing decisions. The opposite is seen when they trade more frequently. We also observe that more sophisticated and experienced traders trade relatively smaller positions and exhibit less overconfidence, consistent with these individuals suffering fewer behavioral biases, for which a likely learning effect is observed.