Institutional Herding and Its Impact on Stock Prices
Abstract (Via Missouri)
Using the trades of 776 institutional investors from 1999 to 2004, we examine the existence and impact of short-term institutional herding. We report robust evidence of herding at the weekly frequency using the Lakonishok, Shleifer, and Vishny (1992) measure and the Sias (2004) measure. More importantly, we find that these weekly herds significantly affect the efficiency of security prices. We document strong evidence of return reversals following short-term sell herds and weak evidence of return continuations following short-term buy herds. Our results are consistent with short-term sell herds being motivated by behavioral considerations and driving asset prices away from fundamental values. Alternatively, the absence of return reversals following short-term buy herds suggest that these herds are information based and help impound new information into security prices.
Introduction (via Missouri)
Institutional investing in the United States has increased dramatically over the last twenty-five years. Institutions now account for more than 60% of all equity ownership and an even greater percentage of the trading volume in U.S. markets (Schwartz and Shapiro (1992); Gompers and Metrick (2001); Jones and Lipson (2004); Boehmer and Kelly (2007)). This concentration of ownership and trading activity raises important questions for both academics and practitioners concerning the impact of institutional trading on securities prices. Of particular interest is whether trading activity is correlated among institutions. Anecdotal evidence suggests that institutions herd together, and that these herds significantly impact stock prices. In a recent Business Week article, Margaret Popper states that “[t]he force of the herd pushes stock prices to extremes, producing short-term buying and selling frenzies … the market eventually snaps out of its panic, and … readjusts to where valuations are fair.” 1 Such institutional herds are often viewed as a short-term phenomenon, occurring over a period of days, and more importantly, are believed to temporarily drive security prices away from fundamental values.