Stock Return and Volatility: An empirical study in the Chinese stock market context

Abstract

Frequently observed evidence strengthens the findings about stock returns exhibiting heteroscedasticity and autocorrelation. ARCH effects have been found in security markets along with an abnormal unconditional sampling distribution. In reality, new statistical instruments such EGARCH account for modeling the conditional second moments effectively.

This paper examines stock returns volatility and heteroscedasticity in the Chinese security market context by using the GARCH model. Empirical analysis, based on daily stock return data, clearly rejects the Martingale hypothesis that future changes of daily stock prices in the Chinese markets are independent of past information. Moreover, the assumption of a time-varying risk premium as captured by the GARCH-M specification is verified. The “day of the week effect” for the daily stock returns is confirmed. The information inefficiency could as well be verified.