Herding behavior among security analysts is described as similar behavior by analysts when forecasting main finance ratios of public companies and when giving investment recommendations. This type of behavior can be divided into two categories based on the different driving forces behind the analysts’ herding behavior.. For instance, if one analyst lacks the ability to research and provide recommendations, he or she may follow, or even copy, a famous analyst’s reports. Such action will result in herding behaviorIn this paper, we build undirected weighted networks to study herding behavior among analysts and to analyze the characteristics and the structure of these networks. We then construct a new indicator based on the average degree of nodes and the average weighted clustering coefficient to research the various types of herding behavior. Our findings suggest that every industry has, to a certain degree, herding behavior among analysts. Furthermore, we relate the two types of herding behavior to stock price and find that uninformed herding behavior has a positive effect on market prices, whereas informed herding behavior has a negative effect.
Manuscript profile