A note on Chinese stock market efficiency: fresh evidence from non-linear unit root test

This paper reexamines the efficiency Chinese stock market for the period December 1990 to February 2010 by extending recent work of Qian et al (2008)2 using the nonlinear unit root test developed by Kapetanios, Shin and Snell (2003)3 and Kruse (2010)4. In doing so, the nonlinearity test of Harvey, Leybourne, and Xiao (2008)5 is used to have an insight into the best specification of the model. The nonlinear unit root tests rejects the null hypothesis of unit root, suggesting that Shanghai stock markets is not weak form efficient, which is contrary to the findings of Qian et al (2008) . In addition, the estimated ESTAR models provide strong evidence that the Shanghai stock market is characterized by a slower speed of mean reversion process. This also may explain why Qian et al (2008) failed to reject the null hypothesis of unit root, since TAR models assume instantaneous change in regimes rather than smooth, which is a characteristic of many financial variables