Regime-dependent herding behavior in Asian and Latin American stock markets

Abstract

By using the smooth transition regression (STR) approach, which captures nonlinearity with regime switches, this paper examines investors' herding behavior in Asian and Latin American markets across market states and volatility regimes considering them as the transition variables. While some markets tend to herd with market consensus in high market regime, China, India, Malaysia, Singapore, Argentina and Brazil, in contrast to previous findings, do not show any nonlinearity across market regimes. Investors in most of the markets except Argentina and Brazil herd in high volatility regime. The driving force in investors' herding is the high level of volatility rather than the low returns during the period of market stress. Some of the markets show the presence of adverse herding with both market and volatility regimes. Herding is also prevalent in high volatility regime of VIX. The findings suggest the important role of U.S. market sentiment affecting the herding behavior of investors.