基于马尔科夫机制转换模型的行业相关性研究及投资组合建议

发布时间：2018-03-28 20:00

本文选题：波动性　切入点：相关性　出处：《首都经济贸易大学》2017年硕士论文

【摘要】：近些年来,随着股票市场不断发展,众多关于股市波动性和相关性以及关于金融资产投资组合最优选择的理论也逐渐得到完善与改进。为了研究股票市场波动性及相关性的问题,本文选取了上海证券交易所公布的上证行业指数数据作为研究对象,分别假定残差扰动项服从正态分布、广义误差分布、偏态广义误差分布,对数据建立GARCH(1,1)模型。随后,文章在GARCH模型中引入马尔科夫机制转换过程,建立两状态MS-GARCH模型,通过模型估计结果显示上证行业指数存在着低波动和高波动两种状态,且处于低波动的期望持续时间较长。通过模型的对比发现,GARCH模型残差扰动项服从偏态广义误差分布的拟合效果比其他分布的拟合效果好;MS-GARCH模型对波动性的描述比GARCH模型更好。之后,对上证行业指数正常状态、低波动状态、高波动状态分别进行了动态相关系数的计算,以上证能源为例,结果表明上证能源与其他九个行业呈现较高的正相关性,而且在2015年至2016年之间的正相关性最高;经过状态转换后的两状态相关性有着明显降低。最后对相关性最高时间段内的数据建立均值-方差模型,通过计算得到最优前沿,发现通过增减上证电信的投资比例能直接影响到投资组合的收益。综上所述,我国的上证行业指数具有高波动性和高相关性的特点,因此在做投资组合最优选择时应密切关注收益率变化和股票市场的相关信息。
[Abstract]:In recent years, with the development of the stock market, Many theories about volatility and correlation of stock market and optimal choice of financial asset portfolio have been perfected and improved gradually in order to study the problem of volatility and correlation in stock market. In this paper, the Shanghai Stock Exchange industry index data published by the Shanghai Stock Exchange are selected as the research objects. The residual disturbance terms are assumed to follow normal distribution, generalized error distribution and skewness generalized error distribution respectively. In this paper, the Markov mechanism transformation process is introduced into the GARCH model, and a two-state MS-GARCH model is established. The results of the model estimation show that there are two states of low volatility and high volatility in the index of Shanghai Stock Exchange. Through the comparison of the models, it is found that the fitting effect of GARCH model is better than that of other distributions, and that the MS-GARCH model describes the volatility better than the other distributions. The GARCH model is better. After that, The dynamic correlation coefficient is calculated for the normal state, low fluctuation state and high fluctuation state of Shanghai Stock Exchange industry index. Taking Shanghai Stock Exchange Energy as an example, the results show that Shanghai Stock Exchange Energy has a high positive correlation with other nine industries. Moreover, the positive correlation between 2015 and 2016 is the highest; after the state transition, the correlation between the two states is significantly reduced. Finally, the mean-variance model is established for the data in the period of the highest correlation, and the optimal frontier is obtained by calculation. It is found that the increase or decrease of the investment ratio of Shanghai Telecom can directly affect the return of the investment portfolio. To sum up, the Shanghai Stock Exchange industry index in China has the characteristics of high volatility and high correlation. Therefore, we should pay close attention to the change of return rate and the relevant information of stock market.
【学位授予单位】：首都经济贸易大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F224;F832.51

【参考文献】