Households Learning in the Dark: New Evidence from Retail Traders

Abstract

This paper uses a new dataset and a specific feature of the French stock exchange to study the behavior of individual investors actively trading stocks. Consistent with prior literature, the trading activity of individual investors increases (decreases) following high (low) performance. I carefully split individual investors' performance into (i) performance related to the exposure of their trades to systematic risk factors such as the market premium, the excess returns on small stocks and the excess returns on value stocks, and (ii) residual performance. I show that trading activity reacts to both (i) and (ii) and that this does not seem to be explained by any strategic behavior. Instead, a simple model where individual investors are ex ante uncertain about their ability and factor exposure and trade to learn about them generates predictions consistent with the above evidence as well as additional predictions about learning dynamics that are also borne out by the data.

Keywords: Learning, individual investor behavior, individual investor performance, household finance

JEL Classification: D14, G11