Regime-switching herd behavior: Novel evidence from the Chinese A-share market

We examine time-variations of herd behavior by proposing a Markov regime-switching model. Our model can not only infer the hidden market state which drives the time-varying herd behavior, but can also capture the empirical characteristics. We conduct a comprehensive empirical analysis of the Chinese A-share market. We find evidence that herding is prominent in volatile regimes, while adverse herding is prevalent during tranquil regimes. Moreover, we conduct a simulation example to explain why previous studies on herding presented conflicting results. Finally, we check for herding effects at factor and industry levels, and employ multiple testing to integrate all the results.