An Empirical Study On The Influence Of Stock Size On Calendar Effect In China's A Share Market

In 1970,Eugene Fama proposed the Effective Market Hypothesis(EMH)theory,arguing that the stock price fully reflects all available information on the market,and that such a market is an efficient market.Under the efficient market hypothesis,all the factors that affect the stock price have been reflected in the stock price,no matter what conditions the investors choose which securities or portfolio of securities are impossible to obtain the yield beyond the market.The efficient market hypothesis has been challenged since its birth,especially since market anomalies such as scale effect and calendar effect were discovered.In developed capitalist markets,the existence of these market anomalies has mostly been confirmed,and some scholars continue to explore the economic explanation behind them.China's securities market began to develop in the 1990s.At that time,market anomalies were one of the hot topics of Western scholars.Therefore,although the A-share market started late,but from the beginning accompanied by scholars on the existence of its anomalies.Because the sample selected in the initial study is very small and the research method is not mature,the conclusions are often not reliable.In recent years,scholars on the A-share market anomaly research enthusiasm has not declined,the sample selection is more and more complete,the use of methods is more reasonable,and then obtained more convincing results.Among them,the research on scale effect and calendar effect has gradually enriched.But looking at the research process of domestic scholars,they always choose a certain period of time or a special stock index when selecting samples,so the empirical results are often very different,or even completely opposite conclusions.Therefore,the study of market anomalies still does not have a unified conclusion of the whole sample.The study of market anomalies has guiding significance for investors,listed companies and market regulators:it can guide investors to make investment decisions,help listed companies to make profit distribution decisions,and help regulators better understand the market in order to improve the regulatory system to reduce systemic risk.Therefore,it is necessary to conduct research on market anomalies as far as possible.Based on the latest foreign research perspectives on the correlation between scale effect and calendar effect,this paper empirically examines the impact of individual stock size on calendar effect,and makes regression analysis on the factors that may have strong explanatory power to calendar effect.The conclusions of this study can not only make up for the deficiencies of the current research results,but also provide inspiration for the internal interaction of other anomalies through the study of the correlation between the two effects.This paper selects a full sample of A shares from January 1994 to December 31,2017,including weekly daily return data,monthly data and quarterly data.After screening the data,they are divided into 5 portfolios according to the total market capitalization at the end of the year.In the grouping,we adopt the method of rolling sample grouping,and divide it into groups at the end of each year to adjust the stock allocation of the portfolio.The main research methods and contents used in this paper are as follows:Firstly,the impact of stock size on calendar effect is empirically studied.This stage of research is divided into three levels,namely,the impact of stock size on the weekly calendar effect,the monthly calendar effect and the quarterly effect empirical.Firstly,we test the stationarity of daily return data,and then do ARCH test on the three levels of return data.For the combination of ARCH test,we use GARCH model to regression,and the combination of ARCH test and the least square method to regression.In addition,the scale premium factor is introduced to further test the significance of the difference of returns caused by scale in different time periods.Secondly,Fama-MacBeth regression is used to analyze the explanatory power of macro-factors for calendar effect excess returns.The macro-risk factors selected in this stage include macro-economic growth rate,inflation rate,term spread,short-term interest rate and money supply.Thirdly,Fama-MacBeth regression is used to analyze the explanatory power of the characteristic risk factors on the three levels of calendar effect excess returns.In this stage,the selected characteristic risk factors are total market value,beta,book-to-market ratio.Financial leverage and turnover.Finally,a theoretical analysis is made on the causes of the calendar effect and the influence of individual stock size on the calendar effect.Through the above empirical tests,the main conclusions of this paper are as follows:(1)The weekly calendar effect of China's A-share market is small company Monday effect;(2)The monthly calendar effect of China's A-share market is small company February-March effect;(3)The quarterly effect of China's A-share market is small company first quarter effect;(4)The calendar effect.The correlation with macro risk factors is not significant,but it can be explained by the characteristic risk factors such as total market value.