Herding and Contrarianism: A Matter of Preference?

Abstract Herding and contrarian strategies produce informational inefficiencies when investors ignore private information, instead following or bucking past trends. In a simple market model, I show theoretically that investors with prospect theory preferences generically follow herding or contrarian strategies, but do so because of future returns as opposed to past trends. I conduct a laboratory experiment to test the theory and obtain an estimate of the distribution of preferences in the subject population. I find that approximately 70% of subjects have preferences that induce herding. Using the preference estimates, I quantify informational efficiencies and predict trade behavior in more general environments.