Home | Business News | Browse by Publication | I | International Advances in Economic Research

The effect of annual earnings announcements on the Chinese stock markets.

Article Excerpt
Abstract

This paper examines the annual earnings announcement effect of the stock markets in China. The investigation is based on events analysis and carried out by modeling the daily changes of stock returns using the M-EGARCH approach, by testing the news effects of annual earnings announcement on the conditional mean of abnormal return and the variance of the returns. It is found that a higher than expected earnings announcement leads to a rise in the conditional mean of stock returns on days before the news announcement and a fall afterwards. The conditional volatility of the changes are significantly reduced by bigger absolute values of reported earnings before the news announcement and increased afterwards, supporting the rejection of semi-strong-form efficiency. (JEL G10, G12, G14)

Introduction

Many studies on the semi-strong-form efficiency of stock market are focused on the analysis of the information content of annual earnings and dividend announcements. The purpose of these public disclosure announcements is to provide information that meets investors' needs for decision-making. According to Fama [1991], in a sub-efficient market, the share price may fail to fully reflect all relevant information, and abnormal returns may be obtained by taking advantage of public information because there is a significant time lag between announcement and full incorporation of the information.

Previous empirical studies on annual earning announcements in the Chinese stock market largely concentrate on the drift effect on the overall market attributed to information disclosure [Zhao, 1998; Chen and Liu, 1999; Chen and Yang, 1999; Chen and Chen, 2002], without considering the precise quantitative relationship between yield change and earnings change given new information disclosure. We investigate these effects using the M-EGARCH model. We intend to discover the precise quantitative relationship between the earnings and the yield shift. Specifically, the daily changes of stock prices are modeled by M-EGARCH to ascertain the existence and the nature of the annual earning announcement effects on the conditional mean and variance of the changes. In the European and the U.S. markets, most such news announcements affect stock prices on the days of the information disclosure and the effect would not continue after the day of disclosure. This means that the European and American markets are basically semi-strong-form efficient stock markets. In contrast, we will show that the Chinese markets are not as efficient as the western markets.

This paper will find evidence of the sub-efficient character of Chinese markets by means of an EGARCH model. The EGARCH model is based on exponential generalised autoregressive conditional heteroskedastic model, which was first put forward by Nelson [1991]. This model is more apt to explain the asymmetric changes of the conditional deviation, and the extent of yield change due to information disclosure. The estimated coefficients of the model would provide us with answers to our questions. This type of modeling is important in our research because of the following reasons.

First, the model used here is based on the conditional heteroskedasticity property of the daily changes, so that the results of empirical study are more coincident with actual condition. Many researchers of Chinese stock markets using other models have often ignored the adverse consequences of heteroskedasticity.

Second, the model will estimate the exact value of the mean of daily changes of stock prices in the events window, and the elasticity relating to EPS (earning per share) over time. This empirical result describes more exactly the effect of annual earnings announcement. Other studies, on the other hand, only distinguish the difference between the good and the bad news.

Finally, and more importantly, using the model we can examine and evaluate the effects of announcement news on the conditional volatility of the abnormal return changes in the events window. The exact value of the variance in response to the news, higher volatility due to good news or bad news and the asymmetric reactions towards good and bad news are also offered by the model.

The rest of the paper is organized as follows. The...

View this article FREE - Now for a Limited Time, try Goliath Business News
Free for 3 Days!