Abstract: Price synchronicity has been studied intensively in the literature and it is particularly relevant to the Chinese stock market. Synchronicity means that stock prices tend to move up or down together, which has happened regularly in the past. This price co-movement, especially during market turmoil, can increase systemic risks and cause significant losses to companies and investors. A market with strong price synchronicity is also considered as a less efficient market, and thus damaging the market's further development. Understanding the fundamental causes of stock price synchronicity is therefore important for policymakers and practitioners.
Among all attempts to explain this phenomenon in the existing literature, herding behaviors of analysts are especially interesting aspect to explore. Analysts are very important participants in the stock market in terms of discovering and spreading information in the market. These people often have strong academic and industrial background, and thus able to extract critical information for companies. Their recommendation and research results are frequently used by companies and investors.
In the behavioral finance literature, herding is a typical issue that causes price deviating from fundamentals. It is also found that herding behavior applies to analyst. Given their importance as an information intermediary, herding behavior among analyst can potentially help explain price synchronicity. Extending from the current literature, and to explore how and to what extent herding in analyst drives price co-movement, an empirical investigation on these issues is provided.
To be specific, four testable hypotheses are set. The main hypothesis on whether herding in analyst contributes to the price synchronicity in China is tested. The second to test what the real driving forces for the herding behavior in analyst are, in other words, whether the herding is driven by information or non-information. Following the existing literature, institutional investors, and market sentiment are included into our modelling framework to further explore the role of herding analysts on price synchronicity. It is expected to see, first, the higher institutional investors, the stronger herding impacts; and second, herding role is stronger in bearish market conditions.
Our empirical analysis uses listed A share firms in China from 2010 to 2017. The original data are collected from CSMAR, Wind and RESSET database. Morck et al. (2000) is followed to construct price synchronicity. Analyst's herding index is the key variable of interest. It is constructed according to Gleason and Lee (2003), who use analyst's forecast of future earnings per share (EPS) and whether the analyst make adjustment towards the average value of all analysts.
Through regression analysis and a set of robustness check, the following conclusions are given. First, it is confirmed that herding among analysts can limit their information discovery role, and thus a driving factor of stock price synchronicity. Second, our analysis using a non-star analysts' herding index show that herding among analyst in China is essentially non-information driven. Third, institutional share can reduce synchronicity and potentially dominate the role from analysts. And last, the fourth hypothesis is confirmed that herding analysts have stronger role in pushing price co-movements in bearish market.
While providing interesting addition to the current literature on understanding stock price synchronicity, this paper has also some important policy implications. Regulating analyst and reforming compensation mechanisms of analysts are needed to give stronger incentives for them to provide objective reports. Further reinforce institutional investors' share in certain companies and improving information disclosure can alleviate price synchronicity and avoid negative role played by analysts. Overall, efforts are needed to establish incentive compatible mechanisms for efficient information distribution in the market.

Key words: analyst, herding behavior, stock price synchronicity, dynamic herding index