Herding in financial markets

Abstract

Due to data limitations and the absence of testable, model-based predictions, theory and evidence on herd behavior are only loosely connected. This paper attempts to close this gap in the herding literature. From a theoretical perspective, we use numerical simulations of a herd model to derive new, theory-based predictions for aggregate herding intensity. From an empirical perspective, we employ high-frequency, investor-specific trading data to test the theory-implied impact of information risk and market stress on herding. Confirming model predictions, our results show that herding intensity increases with information risk. In contrast, herding measures estimated for the financial crisis period cannot be explained by the herd model. This suggests that the correlation of trades observed during the crisis is mainly due to the common reaction of investors to new public information and should not be misinterpreted as herd behavior.