Financial Attention and the Disposition Effect

Abstract

Using a novel brokerage dataset covering individual investors' login and stock trading behavior, we investigate the severity of the disposition effect as a function of attention.Our results show that more attentive investors trade less in line with the disposition effect, suggesting a comparative advantage in incorporating information into financial decision making. Furthermore, we find that high attention is related to a stronger tendency to sell moderate losses, as compared to large ones, while low attention increases an investor's likelihood to sell extreme, rather than moderate, profits. These results are in line with the theory of cognitive dissonance and saliency effects.

Decomposing the Disposition Effect

“…This result suggests a novel testable prediction where the effect of attention on the disposition measure depends on the group of investors. Indeed, in line with our prediction for non-professional investors, Dierick et al (2019) find that more attentive retail investors have significantly lower disposition measures. While demonstrating that the disposition measure is not a proper measure of overall "mistakes," this result further shows that variations in its magnitude can identify their incidence since "low error" and "high error" types imply opposite signs of the disposition measure's comparative static with respect to attention.…”

Measuring the disposition effect

“…This logit regression approach has been applied in a dozen empirical papers studying how the DE is affected by some factor (e.g., past returns for Kaustia (2010), investor intelligence in Grinblatt et al (2012) and investor attention in Dierick et al (2019)).…”

“…However, investor monitoring frequency is difficult to capture since connection data are not often available in the datasets. Even when these specific data are available (see Dierick et al (2019)), the connection rate cannot necessarily be interpreted as the monitoring frequency since investors may monitor their positions on other websites and decide to connect to the trading platform only when they want to trade. This issue will specifically be addressed in Section 5 using a simulation.…”

“…Ideally, we should rely on days when a decision (to sell or to keep) is explicitly made, but determining when this decision occurs is quite difficult. Relying on a dataset that provides information about when retail investors connect to the trading platform (as in Dierick et al (2019)) could be an advantage. However, investors have access to alternative websites providing complete information about any asset and they could connect to the trading platform only once they have decided to trade.…”

“…The first one is a logistic regression, which allows the addition of variables and controls on the right-hand side of the equation. In analyzing the relationship between a given variable and the DE, it is natural to include this variable in the model, but it is also necessary to include interaction term(s), as in Dierick et al (2019) or Frydman & Wang (2020). However, with logistic regressions, a choice again has to be made between using a daily frequency or a sale-based frequency.…”