灰色-GARCH混合模型及其在股票指数中的应用

发布时间：2017-12-31 00:38

本文关键词：灰色-GARCH混合模型及其在股票指数中的应用　出处：《西北农林科技大学》2012年硕士论文　论文类型：学位论文

更多相关文章： 灰色-GARCH混合模型 GM模型 GARCH模型

【摘要】：改革开放以来，中国的金融证券市场得到了很好的和完善的发展，无论男女老少越来越多的国民参与进来，形成了人人参与股票投资的热潮，随着股市的跌宕起伏或喜或悲，对于普通的股民，股票市场的风云变幻一直是他们心中既怕又爱的双刃剑，对于投资机构和大股东而言，股票市场的波动是他们规避风险的重要依据，对于市场监管机构来说，股市的波动一直是其市场监管有效性的重要度量和市场政策制定的重要依据。由此可见，波动率的建模和描述一直都是各方关注的焦点和重点，对学者和股市从业者都有极其重要的意义,并且波动率的计算也为VaR数学模型的建立和计算提供了依据和基础。因此，为了描述和刻画市场的波动，本文在时间序列模型其中主要是广义自回归条件异方差模型和灰色模型基础上，认真研究和总结了以前学者和专家的研究成果，提出了新陈代谢灰色-广义自回归条件异方差混合模型，即灰色-GARCH混合模型。以前的研究结果表明，广义自回归条件异方差模型的残差项应会随着时间的变动而受到过去价格波动或信息冲击等灰色不确定性因素的影响，并随之变化，这对于广义自回归条件异方差模型来说是一个很难明确描述和表达的变量，其结果就是直接的影响了广义自回归条件异方差模型对于波动率的刻画和估计。因此本文采用灰色系统理论的以少量数据资料即能建立起不错的预测模型和对灰色不确定性因素的良好描述和预测等良好特性，对广义自回归条件异方差模型内的残差项建立灰色模型，用这两个模型得到灰色-GARCH混合模型，用它来重新描述和估计市场的波动。因为此模型应用到灰色模型和广义自回归条件异方差模型，是这两个模型的有机的结合，具有灰色模型对灰色信息的良好扑捉和广义自回归条件异方差模型对波动率的很好的表达，所以叫做灰色-GARCH混合模型。为了建立灰色-GARCH混合模型，本文首先介绍了时间序列模型和灰色模型的发展和其现在的研究现状，并且对这两类模型的建模步骤和方法进行了比较全面的介绍，在其基础上本文建立了灰色-GARCH混合模型，随后采用道琼斯中国网站数据，运用Eviews软件和Matlab软件，，对选取的道琼斯中国88指数的数据进行了实证分析，结果表明，与广义自回归条件异方差模型相比较，本文所建立的灰色-GARCH混合模型对波动的表达更贴近市场实际，其对波动率的描述和刻画也更加准确，并且对于在此基础上建立的VaR数学模型的准确性提供了可靠的依据和数据保证。
[Abstract]:Since the reform and opening up, China's financial and securities market has been a very good and perfect development, regardless of the men, women and children more and more citizens participate in the formation of everyone's participation in stock investment upsurge. With the ups and downs of the stock market or happy or sad, for the ordinary shareholders, the stock market has been changing in their hearts both afraid and love double-edged sword, for investment institutions and major shareholders. The volatility of the stock market is an important basis for them to avoid risks. For the market regulators, the volatility of the stock market has always been an important measure of the effectiveness of their market regulation and an important basis for the formulation of market policies. The modeling and description of volatility has always been the focus and focus of all parties concerned, which is of great significance to scholars and stock market practitioners. The calculation of volatility also provides the basis for the establishment and calculation of VaR mathematical model. Therefore, in order to describe and characterize the volatility of the market, this paper based on the time series model, mainly generalized autoregressive conditional heteroscedasticity model and grey model. This paper studies and summarizes the research results of previous scholars and experts, and puts forward the mixed model of metabolism gray and generalized autoregressive conditional heteroscedasticity, that is, the grey GARCH mixed model. The residual term of generalized autoregressive conditional heteroscedasticity model should be affected by grey uncertainty such as price fluctuation or information shock with time change. For the generalized autoregressive conditional heteroscedasticity model, it is difficult to describe and express the variables clearly. The result is that it directly affects the characterization and estimation of volatility in generalized autoregressive conditional heteroscedasticity model. Therefore, the grey system theory can be used to establish a good prediction model and grey model with a small amount of data. Good description and prediction of color uncertainty. The grey model is established for the residual terms in the generalized autoregressive conditional heteroscedasticity model and the grey GARCH mixed model is obtained by using these two models. It is used to redescribe and estimate the volatility of the market because the application of this model to the grey model and the generalized autoregressive conditional heteroscedasticity model is an organic combination of the two models. The grey GARCH mixed model is called the grey GARCH mixed model because of the good capture of grey information by grey model and the good representation of volatility by generalized autoregressive conditional heteroscedasticity model. In order to establish the grey GARCH mixed model, this paper firstly introduces the development of the time series model and the grey model and its present research status. And the modeling steps and methods of these two models are introduced comprehensively. On the basis of these models, the grey GARCH mixed model is established, and then the Dow Jones website data is used. Using Eviews software and Matlab software, the data of Dow Jones China 88 index are analyzed. The results show that the data are compared with the generalized autoregressive conditional heteroscedasticity model. The grey GARCH hybrid model presented in this paper is more close to the market reality and its description and characterization of volatility is more accurate. It also provides a reliable basis and data guarantee for the accuracy of the VaR mathematical model established on this basis.
【学位授予单位】：西北农林科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F224;F832.51

【参考文献】