我国股票市场板块效应实证研究

发布时间：2018-11-19 22:15

【摘要】：在一定的时期内,某一板块内的股票出现了齐涨齐跌的现象,称之为“板块效应”。随着我国证券市场的不断完善、发展和股票种类的增加,各类股票板块的划分越来越明显,“板块效应”日益凸显,从而使投资者根据各股票板块的变化规律来决定投资决策变得越来越具有可行性和收益性。本文研究的目的是通过建立金融计量模型来识别我国股票市场的板块效应,选取的研究样本是银行业板块,同时选取上证综合指数作为市场组合。不合理的度量方法会导致表面上的“异常”现象,所以对市场有效性及“异常”现象的度量需要与预测“预期正常收益”的模型一起来检验,以避免“不良”模型。因此我们先对银行组合的日对数收益率和上证综合指数的日对数收益率做散点图,由散点图看出它们之间的线性关系,我们选定一元线性回归模型作为理论模型。由t检验统计量知回归方程通过了显著性检验,因此,此一元线性回归模型是有统计学意义的。利用银行组合日对数收益率对市场日对数收益率进行回归分析,用经验回归方程来预测“预期正常收益”,由实际值与预期正常收益的差值计算出银行组合的超额收益率。用U检验法我们得知,银行组合的预测正常日对数收益率服从正态分布,它的标准差记为δ,因此实际值和预测值的差值(即超额收益率)大于δ时,才是出现“异常”现象,即出现了板块效应。其次在板块效应存在的基础上,引入相对对数收益率来度量板块效应的强弱,当天超额收益率的绝对值与银行组合相对对数收益率标准差的比值,即为板块效应的强度。实证结果表明：在选取的样本中,股票分别在2010年12月13日、12月15日、12月21日、12月23日、12月24日、12月28日、12月31日、1月7日、1月11日、1月25日、27日、28日、31日、2月16日出现了板块效应,且分别在2010年12月23日、24日和2011年1月27日、28日、31日持续出现板块效应。板块效应的强度依次为1.73、1.12、1.198、2.13、1.85、1.17、1.398、3.32、1.46、1.43、1.14、1.36、1.17、1.02。因此在定性分析和实证的基础上,通过对数据的分析建立起的相应的市场模型对板块效应进行实证研究,可使投资者根据各股票板块的变化规律来决定投资决策,以提高股市的回报率。
[Abstract]:In a certain period of time, the stock in a certain plate appeared the phenomenon of rising and falling, called "plate effect". With the continuous improvement of China's securities market, the development and the increase of stock types, the division of all kinds of stock plates is becoming more and more obvious, and the "plate effect" is becoming increasingly prominent. Therefore, it is more and more feasible and profitable for investors to decide the investment decision according to the changing law of each stock plate. The purpose of this paper is to establish a financial measurement model to identify the plate effect in China's stock market. The sample is the banking sector and the Shanghai Composite Index is chosen as the market combination. Unreasonable measurement method will lead to the appearance of "abnormal" phenomenon, so the measurement of market efficiency and "abnormal" phenomenon should be tested together with the model of "expected normal return" to avoid the "bad" model. So we first make scattered plot of the daily logarithmic rate of return of bank portfolio and the daily logarithmic rate of return of Shanghai Composite Index. We choose the linear regression model of one variable as the theoretical model from the scattered plot to see the linear relationship between them. The statistical regression equation of t test has passed the significance test, so the linear regression model is statistically significant. By using the daily logarithmic return of bank portfolio, the market daily logarithmic rate of return is regressed, and the expected normal return is predicted by empirical regression equation. The excess return rate of bank portfolio is calculated from the difference between the actual value and the expected normal return. Using the U test, we know that the normal daily logarithmic return rate of the bank portfolio is normal distribution, and its standard deviation is 未. Therefore, when the difference between the actual value and the predicted value (that is, the excess return rate) is greater than 未, the phenomenon of "abnormal" will occur only when the difference between the actual value and the predicted value is greater than 未. There is a plate effect. Secondly, based on the existence of plate effect, the relative logarithmic rate of return is introduced to measure the strength of plate effect. The ratio of absolute value of excess yield to the standard deviation of relative logarithmic return rate of bank portfolio is the intensity of plate effect. The empirical results show that in the selected samples, the stocks are on December 13, December 15, December 21, December 23, December 24, December 28, December 31, January 7, January 11, January 25, and January 27, respectively. The plate effect appeared on December 28, 31 and February 16, and continued on December 23, 24, and January 27, 28, 31, 2011, respectively. The intensity of plate effect is 1.73 / 1.12 / 1.198 / 2.131.85 / 1.17 / 1.398'3.32'/ 1.46/ 1.43 / 1.14 / 1.36 / 1.171.17 / 1.02. Therefore, on the basis of qualitative analysis and empirical analysis, through the corresponding market model established by the analysis of the data, the empirical research on the plate effect can make investors decide the investment decision according to the changing law of each stock plate. In order to improve the return on the stock market.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;O212.1

【参考文献】