Does Asymmetric Information Drive Herding? An Empirical Analysis

Abstract

The authors explore the relative significance of information and its role in herding formation among investors. They investigate this relationship at 3 scales: the market, investor, and individual stock levels. At the aggregate market level, empirical evidence obtained suggests that a more transparent environment is likely to mitigate the extent of herding intensity, mainly as a result of a decay in non–information-based “intentional” herding. At the investor level, the authors find strong evidence of asymmetric herding between investors with heterogenous information (arbitragers and noise traders). Finally, at the stock individual level, the present results show a positive and statistically significant relationship between herding intensity and 5 firm-specific measures of information asymmetry.

Keywords:

• Chinese equity markets
• Herding behavior
• Information asymmetry
• Market liberalization

Notes

Notes

1 Chang et al. [2000] used similar figures to demonstrate the relationship between the cross-sectional absolute deviation (${\mathit{CSAD}}_t)$ and equal-weighted market return ($R_{m,\mathrm{ }t}$). In their study, they focused on the presence of a nonlinear relationship between ${\mathit{CSAD}}_t$ and $R_{m,\mathrm{ }\mathrm{ }t},$ as this implies the presence of herding (see Chang et al. [2000] for more information on ${\mathit{CSAD}}_t$). Similar to their study, we present the relationship between ${\mathit{CSAD}}_t$ and $R_{m,\mathrm{ }\mathrm{ }t}$ in the upper and lower tails, and before and after market liberalization.

2 We considered 2 additional proxies of information asymmetry: probability of informed trading and adjusted probability of informed trading. While data needed to calculate both measures is available over the postliberalization period, it is not over the preliberalization period.