Institutional Herding: Causality and Persistence

We investigate the causality in herding between foreign portfolio investors (FPIs) and domestic mutual funds (MFs) in the Indian stock market. The estimated herding levels are considerably higher than those observed in other international markets, and herding is prevalent in small stocks. We find that institutional investors follow contrarian-trading strategies, unlike what was documented in most other markets. Analysis of the aggregate herding measure shows a bi-directional causality between FPIs and MFs. Further analysis using directional herding measures indicate no evidence of causality between institutional herds on the sell-side. But we find causality on the buy-side and it is running in both directions between FPIs and MFs, implying a feedback of information. Given the tendency of institutions for herding in small stocks, adopting contrarian-trading strategies, the observed sell-side causality is perhaps having a salubrious effect. As institutional investors are contrarians, their trading activity will lead to price corrections in small stocks aligning with the fundamentals, thereby contributing to market efficiency. JEL Classification: C23, C58, G23, G15, G40

Herding on Earnings News: The Role of Institutional Investors in Post–Earnings-Announcement Drift

We examine the role of institutional investors underlying post–earnings-announcement drift (PEAD). Our results show that while institutional investors generally herd on earnings news, such correlated trading among institutions does not eliminate or reduce market underreaction to earnings surprises. Instead, PEAD is significant only in the subsample of stocks where institutions herd in the same direction as earnings surprises. In fact, institutional herding is also positively related to next-quarter earnings announcement returns. We provide evidence that institutional herding on or against earnings news is largely driven by firm characteristics, particularly past firm performance and stock returns. In addition, we find that relative to nontransient institutions, transient institutions have a stronger tendency to herd on earnings information. Finally, based on long-run stock returns, we show that when institutions herd on earnings surprises, institutional trading represents a gradual process of incorporating information into stock prices. However, when institutions herd against earnings surprises, institutional trading slows down stock price discovery.