A Study On Day-of-the-week-effect In China's Stock Mark

Efficient market theory is one of the cornerstones of finance, but in the real economy, there are many phenomena challenging this cornerstone, one of which is the Day-of-the-week-effect. Day-of-the-week-effect refers to the return of a particular trading day significantly higher or lower than the other trading days, and this phenomenon persistently exists. According to efficient market theory, investors are rational, the return differences of trading day will arise no risk arbitrage opportunity, rational investors' arbitrage behavior will lead to market equilibrium and then eliminate these differences. However, empirical evidences indicate that Day-of-the-week-effect is widely exists in financial markets.After reviewed the relative literature, we found that the existing literature, in particular, the domestic literature deserved to be deepened in two aspects. First, for the research instrument, majority of existing studies either use linear regressive model or the GARCH series model, whereas non-normal distribution characteristics of return, as well as the statistical test lack of financial economics foundation makes the analysis based on these tools to be challenged. Second. for sample selection, different researcher chooses different market indexes, although the same market index, but date span is different, thus the results can not distinguish whether the Day-of-the-week-effect actually exist or data mining caused. This paper intends to perfect above two aspects.In our paper, we apply Stochastic Dominance to analysis the Day-of-the-week-effect in China's stock markets. Compared to GARCH and regressive models, stochastic dominance method has two advantages. First, it is distribution free, so we don't require normal distribution of return. Second, about investor's utility function, stochastic dominance criterion's assumptions about the utility function meet the majority of investors in stock markets. In the data Selection, in order to rule out the historical data accumulate in the sample and cover up the economic information contained in new data, in this paper sample rolling method is used. Sample rolling method can make us to test the stability of Day-of-the-week-effect as time going. Besides, we make comparative analysis about different indexes to test whether index selection affects the research results.After analysis, we have some conclusion. First, the Day-of-the-week-effect exists in China's stock markets. high return trading day appears between Wednesday and Friday, the low return trading days appears between Thursday and Tuesday. Second, the return's dominant relation of most trading days can be judged under second order, so we can regard the analysis result as the decision of risk aversion investors in stock market. Third, in comparative analysis, we have the same conclusion for different index. Last, through sample rolling analysis, Day-of-the-week-effect manifests very stable, but the index daily return distribution within the week manifests a dynamic characteristics.