Offsetting the disposition effect with a stop-loss rule

Abstract

We put forward an agent-based model of the stock market, where the behavior of agents showing the disposition effect can be offset by that of others using a stop-loss rule. In a stop-loss order, a stock is sold automatically if it drops below a threshold value. The disposition effect is the tendency to sell stocks that have gained in value (“winners”) and keep the ones that have fallen in value (“losers”). After showing the model can replicate actual return behavior considering data from the recent mini flash crashes, we explore the consequences of altering key behavioral parameters. Our primary result is that the presence of stop-loss agents in a non-Gaussian environment can offset the disposition effect. Furthermore, we find differing return targets to contribute to market efficiency, and a negative shock to a market sentiment index to cause the stock price to dip and trade volume to grow. Finally, increasing overconfidence generates higher trade volume.