The influence of the US market on herding behaviour in China

Abstract

This article investigates the influence of the US market on the herding behaviour on Chinese financial market through an analysis of daily data from the Shanghai and Shenzhen stock exchange markets for the period 2006–2012. This period is very informative because the financial crisis that emerged on the US market quickly widespread at a global level and that specific situation can generate herding behaviour. Results confirm the influence of the US market on the Chinese stock markets, but they show there is no contagion effect between these two countries. These results can be partly explained by the difference in terms of market structure: China stock markets have a unique micro- and macro-structure within which the government can easily intervene in case of destabilizing situation while the US markets are mainly independent of government.

Keywords:

• herding behaviour
• Chinese stock markets
• influence of the US market

JEL Classification:

• G15
• G14

Notes

¹ These companies have been carefully selected by using two criteria: the necessity to adjust companies in the same date format meaning that we did not deal with companies which suspended trading and the second criterion referred to the importance of capitalization of firms. Selected companies are the most-capitalized companies on both markets. Data have been collected from China Stock Market Trading Database (CSMAR) (http://www.gtadata.com/) and Bloomberg Database.

² The choice of working with daily returns is consistent with that of Christie and Huang (1995) or Tan et al. (2008) who showed that herd formation is short-lived behaviours. The daily return is calculated by the equation: *100, where denotes the closing price of firm i on time t.

³ In order to calculate the return dispersion and weighed average of US market, top 30 largest market capitalization listed companies were chosen.

⁴ We tested the robustness of data by doing the autocorrelation test of residuals and the unit root test which showed both stationary variables and no residual autocorrelation. We also added a lagged-dependent variable into the equation in order to detect potential herding behaviours, and the improved model did not show a significant coefficient and therefore confirms that US financial market has no important impact on herding behaviour in China.