Recent empirical studies have documented a number of regularities in the behavior of investors that seem to be at odds with the rational expectations paradigm. One of the most striking patterns is the tendency of investors to sell their winners and to hold on to their losers. Such behavior, which has been termed the “disposition effect” by Shefrin and Statman (1985), has been found in a variety of data sets and time periods.
Although the existence of the disposition effect seems undisputed, economists and investment professionals have not agreed on an explanation for this phenomenon. The empirical literature favors a behavioral explanation offered by Shefrin and Statman (1985), which combines the ideas of mental accounting (Thaler (1985)) and prospect theory (Kahneman and Tversky (1979)). Shefrin and Statman argue that investors keep a separate mental account for each stock. Within that account, investors maximize an “S”-shaped valuation function, which is similar to a standard utility function except that it is defined on gains and losses relative to a reference point (usually the purchase price), rather than on absolute wealth. This valuation function is concave in the gains region and convex in the loss region. Thus, if a stock appreciates in price, the investor’s wealth will be in a more risk-averse part of her valuation function, making a sale more likely. In contrast, if the stock is trading below its purchase price, the investor becomes risk-loving, and will hold on to the stock for a chance to break even.
In addition, there are rational explanations for the disposition effect. First, portfolio rebalancing considerations suggest that investors who do not hold the market portfolio should respond to large price increases by selling some of the shares they hold in these stocks to restore diversification (Lakon-ishok and Smidt (1986)). Second, since transaction costs tend to be higher for lower priced stocks, and since losing investments are more likely to be lower priced, investors may refrain from selling losing investments simply to avoid the higher transaction costs (Harris (1988)). Finally, disposition behavior may result from informational differences across investors (Lakonishok and Smidt (1986)). An investor who purchased a stock on favorable information may sell it when the price goes up because she rationally believes that the stock price now reflects this information. On the other hand, if the price goes down, the investor may continue to hold it, rationally believing that her information has not yet been incorporated into the price.
These alternative rational explanations have been challenged by recent empirical studies. Odean (1998) shows that investors who sell their entire holdings of a stock — and who are thus unlikely to be motivated by diversification — continue to prefer selling winners. In addition, he provides evidence against the hypothesis that higher trading costs for lower priced stocks are responsible for the disposition effect. Even when differences in transaction costs are controlled for, investors appear to be reluctant to realize their losses. Moreover, Odean (1998) and Brown et al. (2002) argue that the investors’ preference for realizing winners rather than losers does not appear to be justified by the subsequent stock performance. Both studies find that, on average, winners that are sold outperform, over the subsequent six to 24 months, losers that are not sold, which leads them to reject the information-based explanation suggested by Lakonishok and Smidt.
This paper suggests that the case for rational disposition effects is not hopeless, however. In fact, a simple rational expectations model with asymmetrically informed agents can generate most of the empirically documented facts. The aim of this paper is threefold. First, we demonstrate that, depending on how the degree of information asymmetry changes over time, not only better informed investors choose to sell their winners rather than their losers (as suggested by Lakonishok and Smidt), but also less informed traders might prefer such a strategy. Second, we show that risk-averse investors can rationally exhibit disposition effects even though past winners continue to outperform past losers in subsequent periods. In other words, our results indicate that the existing empirical tests rejecting an information-based explanation for the disposition effect are inconclusive. Third, we examine the relationship between the investors’ trading behavior and equilibrium price dynamics. In particular, we show that both a disposition effect and price momentum can arise in a world with fully rational agents.
Download
Information Asymmetry, Price Momentum, and the Disposition Effect
