The informational content of earnings announcement and stock market behaviour : an investigation of China's stock market

Abstract

The purpose of this research, is to empirically examine the information content of earnings announcements and price and volume movements in the emerging Chinese stock market Specifically, the study attempts to 1) investigate whether earnings announcements have significant effects on security returns and trading volume; 2) describe factors, including some unique to China, involving information content and their effects on price/volume reactions to earnings announcements; 3) examine the timeliness of annual report releases and market reaction to earnings announcements in the Chinese stock market; and 4) study the interaction effect of contemporaneous announcements of earnings and dividends on share price. First of all, we adjust the original data based on the Center for Research in Security Prices (CRSP) standard to make it comparable. Then, both, the Random Walk Model and the Growth Model are introduced to estimate unexpected returns, and the Market Model is introduced to estimate market reactions. In addition, we choose four different event windows to capture the cumulative announcements' effects, and four sub-samples to make comparisons. Based on a set of 1,499 earnings announcements, non-parametric statistics as well as multivariate regressions are employed. To measure the relationship between information content surrogates and price/volume reactions to earnings announcements, we hypothesize that cumulative abnonnal returns are positively related to unexpected earnings and public ownership, and negatively to firm size, while the absolute value of cumulative abnonnal returns are positively related to the magnitude of unexpected earnings and public ownership, and negatively to firm size. We also hypothesize that cumulative abnormal trading volume is positively related to the magnitude of unexpected earnings and public ownership, and again negatively related to firm size. As to the relationship between volume and price reaction, we hypothesize that there exists a positive correlation between abnormal trading volume and abnormal return, and between the former and the magnitude of the latter. Moreover, abnormal trading volume is higher for price increases than for price decreases. The test results support most of our hypotheses. As for the timeliness of annual report releases and market reaction to earnings announcements, we hypothesize that firms with favorable private information would reduce information asymmetry through an earlier announcement. Furthermore, firms with unfavorable private information would intentionally maintain the asymmetry by a later earnings announcement Our results indicate that in China, annual earnings announcements lead to significant volume reactions. In addition, companies announcing their annual earnings in earlier months received significantly stronger abnormal market reactions than those making announcements later in the year. We argur that selecting the timing of earnings announcements reflects the type of information, as well as the information asymmetry of the companies. Firms willing to make early announcements tend to possess favorable private information and such announcements surprise the market by a larger extent. Those making late announcements tend to possess negative private information and these later announcements are more easily predicted by the market. Our findings lend strong support to this hypothesis of trading volume reaction. In addition, we evaluate the price effect of concurrently announced earnings and dividends because they are almost always announced simultaneously in China. Investors might want to verify the news content of one signal with the other. The interaction is tested by employing an 'interaction' F-statistic and a 'first-order' F-statistic as used by Kane et al. (1984) and Easton (1991). The null hypothesis is that there are no interaction effects between earnings and dividend announcements. Our result is consistent with previous evidence; that earnings and dividends announcements have information content and can, in themselves, induce share price adjustments, and there exists a statistically significant interaction effect. However, we find no systematic patterns of combined earnings and dividend effect. The presence of a corroborative effect cannot be detected. We do not observe a larger magnitude of abnormal returns when these two variables change in the same direction, as compared to when they change indifferent directions. Our results also indicate that the market does not care much for collaborating evidence from the dividend signal. Moreover, to make our results more robust, we conduct an additional test on market reaction to earnings announcements after controlling rights offering in the Appendix. The results are very similar to the results without controlling rights offering. This research should complement and strengthen the current understanding of the Chinese stock market's efficiency and related issues. It is one of the first attempts to extend the western research framework regarding event studies to the rapidly emerging Chinese stock market It is also helpful for measuring information content and the market behavior of earnings announcements.