Literature on financial economics suggests that market participants tend to suppress their own information and try to imitate others in the market, thereby herding against their private information. This tendency is attributed to the risk aversion characteristic of economic agents who rely more on short cuts and heuristics in order to avoid the risk of losing time required to incorporate private information. Such a tendency also results in the asymmetric expected returns on assets. We attempt to find empirical evidence of herding in two different cross-sections of financial markets using cross-sectional deviations of stock returns to measure the dispersion of individual stock returns from average market return. Using a unique dataset of daily stock returns from January 2011 to December 2015, we examine the small- and large-cap stocks for the effect of herding. We study the existence of herding in two cross-sections of stocks in the Indian stock market, and show that stocks with robust fundamentals observe little or negligible evidence of herding while vulnerable stocks are evidently found to be affected by herding. While examining herding, we see if the cross-sectional dispersion of stock returns in large-cap stocks are lower as compared to that in small-cap stocks, implying that stocks with higher market capitalisation and trading volume are less prone to herding. We conclude that among all measures of herding in stock markets, cross-sectional absolute deviation is a reasonably good proxy. Our empirical results ascertain the existence of herding more particularly in large-cap stocks traded on the National Stock Exchange of India. As highlighted in prior literature, market-wide aggregate herding is a priced risk factor for large-cap stocks, but not for small-cap stocks. In an emerging market such as India, herding among investors is attributed to information asymmetry. The evidence of the cross-sectional variation contributes to the existing literature on investor behaviour and its implications for asset pricing. An extension of this work could be a cross-market analysis of herding for a wider selection of stocks.
