||A comment investment question is: Is herding behavior rational or irrational? Previous studies don’t propose effective methodologies to identify this behavior, or deeply investigate the differences of both rational and irrational herding behavior. The purpose of this study is to propose some methodologies to measure rational herding. Initially, we use the “S statistic” proposed by Bernhardt, et al. (2006) which was originally used to estimate the probability of analysts’ herding and apply it to the stock market. Second, we use the condition of following return to identify order imbalance as rational herding. Then, the residuals form the regression of that unexplained by rational herding. Finally, we research the interaction between rational and irrational herding and market volatility, efficiency, and noise using the Vector Autoregression Model. |
The findings of this paper are: (1) Rational herding can decrease volatility and noise, but has no significant impact on efficiency in high market value situations. (2) Rational herding can also decrease noise but increase volatility, however there is also no significant impact on efficiency in low market value situations. (3) Irrational herding can increase volatility and noise but has no significant impact on efficiency in either high or low market value situations. (4) Volatility and noise can decrease rational herding in high market value situations, but the volatility increases rational herding and noise still decreases rational herding in low market value situations. (5) Volatility and noise can increase irrational herding in both high and low market values situations. (6) Market efficiency has no impact on rational or irrational herding in either high or low market value situations.