A Dynamic Measure of Intentional Herd Behavior Causing Excess Volatility in U.S. Stock Markets

This paper suggests a dynamic measure of intentional herding, causing the excess volatility or even systemic risk in financial markets, which is based on a new concept of cumulative returns in the same direction as well as the collective behavior of all investors towards the market consensus. Differing from existing measures, the measure allows us to directly detect time-varying and market-wide intentional herding using the model of Dynamic Conditional Correlation (DCC) (Engle, 2002) between the financial market and its components that is partially free of spurious herding due to the inclusion of the variables of the number of economic news announcements as a proxy of market information. Strong evidence in favor of the dynamic measure over the other measures is based on empirical application in the U.S. markets (DJIA and S&P100), supporting the tendency to exhibit time-varying intentional herding. More importantly, it is found that the impact of intentional herding on market volatility tends to be stronger during the periods of turbulent markets like the degradation of U.S. sovereign credit rating by S&P, and be more significant in S&P 100 than DJIA.