“Herd behaviour, volatility clustering and the Indian stock market: An empirical investigation”

This paper examines the phenomenon of herding behaviour in the Indian stock market and its implications for market volatility under conditions of uncertainty. We use daily observations from 2013 to 2024 of the Nifty 500 and apply the cross-sectional absolute deviation approach, the Vector Autoregressive Model, the Exponential Generalized Autoregressive Conditional Heteroskedasticity model and the Markov Regime Switching model to study interrelated dimensions of herding behaviour, market volatility, and the persistence of collective trading patterns in varying market conditions. We find strong evidence of herding behaviour in the Indian stock market, particularly during down markets. In contrast, no herding is observed during up markets. Herding is found to be more intense and persistent during high-volatility periods, while it is weaker and short-lived during low-volatility regimes. The study also concludes that trading volume, market volatility, and return dispersion do not Granger-cause herding. Furthermore, we present evidence of volatility clustering, where both positive and negative shocks increase future volatility, although negative shocks have a stronger impact. The findings of this study have policy implications in terms of the need for stronger awareness and risk-disclosure frameworks for retail investors, as we report that intensified herding occurs during high-volatility and crisis regimes. Similarly, volatility-aware portfolio rules, dynamic liquidity buffers, and counter-cyclical trading strategies can be beneficial in cases of asymmetric responses to negative shocks and the formation of negative feedback loops. Further, based on the results, we discuss the need for market-stabilization tools, such as dynamic circuit breakers, tighter margin requirements during periods of stress, and real-time volatility monitoring.