基于数字文化的中国股票市场日历效应的实证研究

发布时间：2018-01-19 14:35

本文关键词： 数字文化择日效应 4日期效应 GARCH模型　出处：《西南交通大学》2013年硕士论文　论文类型：学位论文

【摘要】：作为备受关注的金融市场异象之一,日历效应被证明普遍存在于不同国家、不同地区的股票市场中,大量的文献研究已对其作了较为深入的探讨。本文希望在前人研究基础上从数字文化视角来探讨我国股票市场是否存在择日效应。选择这个视角,方面是因为数字文化作为中国传统文化的重要内容,对国人的生活有深刻的影响,而目前研究数字文化对股票市场影响的文献并不多；另一方面日期是以数字来表示的,有助于研究工作开展。本文以2000年1月至2012年12月十三年间共3144个交易日的上海、深圳A股市场上证综合指数和深证成份指数每日收盘价作为研究样本,数据来自CCER中国经济研究数据库。同时,还对上市公司召开股东大会及IPO的日期做了简单分析,这部分数据分别来自同花顺和CCER数据库。本文通过建立回归模型和GARCH模型对指数收益率及其波动的日历效应进行检验,对农历日期和公历日期分别研究,并将日期按照尾数分组探讨。回归模型检验结果表明,农历日期只有初三和二十八在10%的水平上显著,但尾数分组检验并不显著。对公历分组数据而言,4日期存在显著为正的收益率,可认为我国股市存在4日期效应；8日期收益并不显著,但7日期通过弱检验存在显著为负的收益,9日期通过强检验存在显著为正的收益,8日期前后的异常收益可以归结为8日期效应。GARCH模型检验结果表明,农历日期不存在显著的收益,公历日期只有4日期表现为显著的正收益,波动则呈现显著为负的3日期效应和5日期效应。研究结果表明,由于公历的普遍运用,投资者交易时更关注公历日期：一方面，，投资者对数字有明显偏好,在公司IPO及召开股东大会时,更多选择8日期而避开4日期：另一方面,我国股市受到数字文化影响而存在择日效应,即4日期存在显著的正收益,且在4日期前后会出现显著的异常波动,本文认为这是数字文化影响了投资者交易心理、交易行为从而表现为市场收益和波动的异常。
[Abstract]:As one of the most concerned anomalies of financial markets, calendar effect has been proved to exist in different countries and different regions of the stock market. A large number of literature studies have made a more in-depth study of it. This paper hopes to explore whether there is a day effect in the stock market from the perspective of digital culture on the basis of previous studies. The reason is that digital culture, as an important part of Chinese traditional culture, has a profound impact on the life of Chinese people, but there are few literatures on the influence of digital culture on stock market. On the other hand, the dates are expressed in numbers, which is helpful to the research work. This paper takes the daily closing price of Shanghai Composite Index and Shenzhen Stock Exchange component Index in Shanghai and Shenzhen A-share market from January 2000 to December 13th 2012 as the research samples. The data are from the CCER China Economic Research Database. At the same time, the date of the shareholders' meeting and the date of the IPO of the listed company are also analyzed briefly. This part of data comes from Tonghuashun and CCER database respectively. This paper tests the calendar effect of index return and its fluctuation by establishing regression model and GARCH model. The dates of lunar calendar and Gregorian calendar are studied, and the dates are divided into groups according to Mantissa. The results of regression model test show that only the third and 28 lunar dates are significant at the level of 10%, but the Mantissa grouping test is not significant, as far as the Gregorian calendar grouping data are concerned. (4) there is a significantly positive rate of return on the date, which can be considered as a 4-day effect in China's stock market; 8. The yield on date is not significant, but there is a significant negative return on the weak test of 7 days. The yield of 9 days through the strong test is significantly positive. The results of the GARCH model test show that there is no significant income on the lunar calendar, and only 4 days in the Gregorian calendar show significant positive returns. The fluctuation showed negative 3-day effect and 5-day effect. The results show that due to the universal use of Gregorian calendar, investors pay more attention to Gregorian calendar dates: on the one hand, investors have a clear preference for numbers, in the company IPO and the convening of shareholders' general meeting. On the other hand, due to the influence of digital culture, China's stock market has a day effect, that is, there is a significant positive return on the fourth day, and there will be significant abnormal fluctuations before and after the 4th day. This paper argues that the digital culture has influenced investors' trading psychology, and the trading behavior is characterized by abnormal market returns and fluctuations.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224

【参考文献】