Conditional Herding in the Indian Stock Market: The Role of Investor Sentiment and Global Contagion

Abstract

This study investigates conditional herding behaviour in the Indian equity market from April 2015 to March 2025, employing a survivor portfolio of 32 consistently traded Nifty 50 constituents. Using the Chang, Cheng, and Khorana (2000) cross-sectional absolute deviation (CSAD) framework with sentiment interaction terms, we document evidence of herding that intensifies during periods of elevated market fear, as measured by India VIX. After controlling for heteroskedasticity and autocorrelation using Newey-West standard errors (lag 5), the interaction coefficient between squared market returns and standardized sentiment is −1.654 (t = −2.33, p = 0.020), indicating that investors reduce fundamental differentiation and converge toward consensus during high-volatility regimes. When controlling for global market spillovers via lagged S&P 500 returns, the sentiment-herding relationship remains statistically significant while the global spillover coefficient becomes statistically insignificant (coefficient = 0.035, t = 1.46, p = 0.145), suggesting herding in India is primarily driven by domestic sentiment rather than international contagion. Robustness checks using the broader Nifty 50 index confirm these findings (interaction coefficient = −0.255, p = 0.014). Our results contribute to understanding behavioural anomalies in emerging markets and have implications for portfolio diversification, regulatory oversight, and market stability.