Does volatility mediate the impact of analyst recommendations on herding in Malaysian stock market?

Abstract This study examines the mediating role of volatility on the relationship between analyst recommendations and herding in the Malaysian stock market by using data from 2010 to 2020. Volatility is measured by realized volatility and the Parkinson estimator. The empirical evidence suggests that herding exists and realized volatility intervenes in the direct relationship between analyst recommendations and herding. The release of analyst recommendations causes realized volatility to fluctuate and investors are triggered by the volatility, which in turn follow the crowd to herd. Nonetheless, the Parkinson estimator is found to be insignificant, which infers that investors have anchor bias and rely on previous day stock prices to trade and herd. This paper provides an alternative explanation to the direct relationship and enhances the study of information-based herding. It contributes to academicians, practitioners, investors and policymakers to understand the herding of investors in responding to the arrival of new information.