In financial markets, herd behavior refers to the process where market participants disregard their private information and instead base their financial decisions upon the actions of other participants, and trade in the same direction. Investors’ herd behavior exacerbates market volatility, generates serial dependence in trading patterns, and causes informational inefficiencies. If many traders herd, the security may have a price that does not necessarily reflect its fundamental value. Such deviations from fundamental values may lead to short-run mispricing like flash crashes but may also cause asset price bubbles and other extreme market outcomes, such as severe financial crises. Hence, understanding how investors’ herding behavior affects financial markets is important for both policymakers and for all financial market participants. The purpose of this project is to provide a new herding measure and then use it to study its impact on stock market efficiency, liquidity, and volatility. There is a disconnection between the empirical and theoretical research on herding in existing literature. Theoretical papers show that in an abstract environment where agents with private information make their decisions in sequence and after a finite number of agents have chosen their actions, all others will disregard their own private information and herd. However, it is difficult to directly apply the theoretical model empirically because private information is unobservable. Therefore, the empirical research on herd behavior identifies herding in financial markets through statistical measures of return clustering. However, the statistical measures cannot distinguish true herding from spurious herding, that is when decision clustering are due to a common reaction to public announcement and not a consequence of investors imitating each other. Cipriani and Guarino (2014, CG14 hereafter) propose a new theoretical framework explaining informational herding that extracts the private information from transaction data (i.e., trade and quotes in real time). This herding measure enables us to distinguish true herding from spurious herding. However, the framework of CG14 simply assumes that the public announcement is irrelevant to investors’ herding. But according to Hirshleifer & Sheng (2022), the arrival of important public news, such as a Central bank meetings (FOMC) meeting, can trigger a shift in investors’ attention towards focusing on the impact of the FOMC announcement on the fundamental value of the stock that they hold, thus affect their private information set on the relevant stock. Recognizing that expected shocks (e.g. public announcements) may cause spurious herding, findings by Hirshleifer & Sheng (2021) imply that unexpected public announcement shocks affect investors’ private signals, resulting in true herding. The project aims to extend CG14’s framwork to identify true herding where we consider not only private information but also the impact due to unexpected public announcment shocks. With this new measure, we study how herding may affect stock market efficiency, liquidity, and volatility.