基于状态转移-Copula模型的股市量价关系研究

发布时间：2018-01-19 01:09

本文关键词： 量价关系马尔科夫 MRS-GARCH模型马尔科夫状态转移混合Copula模型极大似然估计　出处：《西南交通大学》2014年硕士论文　论文类型：学位论文

【摘要】：伴随着经济的飞速发展,经济全球化与金融一体化促使金融各个市场之间的相互依存性大大地增强,市场与金融资产之间的相关关系也变得越来越复杂。中国股市历经二十多年的风雨,已经趋于完善和成熟,其股票的量价之间的关系呈现了非线性、非对称性等模式。本文主要研究了基于状态转移GARCH模型(MRS-GARCH模型)与状态转移混合Copula模型下的中国股市量价关系。利用状态转移来刻画股票市场的价格与成交量的状态变化过程,并选取了GARCH模型和混合Copula模型,然后把它们分别与马尔科夫状态转移相结合的方法,对中国股市的上证指数和房地产板块指数进行相关建模分析。根据股票的价格和成交量所具有的尖峰厚尾特性,对其建立了MRS-GARCH模型,并通过拟合优度检验证明用此模型是合理的。在实证分析中,把GARCH模型与MRS-GARCH模型通过AIC值进行比较,结果表明MRS-GARCH模型对刻画股市的量价关系更好一些,并且由参数估计值知：中国股市中成交量对价格波动具有一定解释作用,且中国股市的上证指数和房地产板块指数出现了不同的波动状态。虽然GARCH模型在实际应用中具有非常多的优点,但是它不能对量价关系的非对称性和尾部结构较好地刻画,在研究量价关系时必须充分考虑两个随机变量间的局部相关结构及差异性。鉴于此,本文利用了状态转移混合Copula模型进行了相关分析,并且利用极大似然估计的方法,对该模型的参数进行了估计,最后通过蒙特卡洛模拟的方法对模型进行了检验,得出了很好的结论。实证结果表明用马尔科夫状态转移混合Copula模型刻画中国股市的量价相关结构是十分合理的,能够捕捉到在不同波动状态下量价之间的非对称、非线性及尾部结构关系,比MRS-GARCH模型具有更大的优越性。
[Abstract]:With the rapid development of economy, economic globalization and financial integration promote the mutual dependence of various financial markets greatly enhanced. The relationship between the market and financial assets has become more and more complex. After more than 20 years of wind and rain, Chinese stock market has become more and more perfect and mature. Asymmetrical isoforms. This paper mainly studies the MRS-GARCH model based on state transition GARCH model. The relationship between volume and price of Chinese stock market under the mixed Copula model of state transition. The state transition is used to describe the state change process of price and trading volume in stock market. The GARCH model and the mixed Copula model are selected, and then they are combined with Markov state transition. Shanghai Stock Exchange Index and Real Estate Plate Index of Chinese stock market are modeled and analyzed. According to the characteristics of stock price and volume, the MRS-GARCH model is established, and the goodness of fit test proves that the model is reasonable. Comparing GARCH model with MRS-GARCH model through AIC, the result shows that MRS-GARCH model is better for describing the relationship between volume and price of stock market. And from the parameter estimate, we know that the trading volume in the Chinese stock market has a certain explanatory effect on the price fluctuation. And the Shanghai stock index and real estate sector index of Chinese stock market have different fluctuation state, although GARCH model has many advantages in practical application. However, it can not describe the asymmetry and tail structure of the relationship between quantity and price. In order to study the relationship of quantity and price, the local correlation structure and the difference between two random variables must be fully considered. In this paper, the state transition mixed Copula model is used to analyze the correlation, and the maximum likelihood estimation method is used to estimate the parameters of the model. Finally, the model is tested by Monte Carlo simulation. The empirical results show that it is very reasonable to use Markov state transition mixed Copula model to describe the volume and price correlation structure of Chinese stock market. It can capture the asymmetric, nonlinear and tail structure relationships between quantities and prices under different fluctuating states, which is superior to the MRS-GARCH model.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F830.91;O211.62

【参考文献】