基于t-Copula-EGARCH模型的沪深股市风险值的研究

发布时间：2018-01-11 15:14

本文关键词：基于t-Copula-EGARCH模型的沪深股市风险值的研究　出处：《西南财经大学》2013年硕士论文　论文类型：学位论文

更多相关文章： VaR t-Copula EGARCH模型 事后检验

【摘要】：随着金融衍生产品与信息技术的发展,全球金融市场的交易量规模越来越大。然而金融市场的欣欣向荣,也潜藏着巨大的风险,如2007年的美国次贷危机、2009年底发生的希腊债务危机等。金融危机爆发给投资者带来了巨大的损失,因此如何有效的计算金融资产的风险价值成为各个金融机构面临的很重要的问题。通常,我们计算金融资产的风险值有三类主要方法：敏感性分析法、历史收益模拟法、蒙特卡洛模拟方法等。单个金融资产的VaR易于计算,而多个金融资产的联合分布下VaR的计算比较困难。以往,我们通常假定多个金融资产服从线性相关性,然后计算出风险值。然而多个金融资产之间并不服从线性相关性,因此本文引入Copula理论来描述多个金融资产的相关性。在实证部分,我们先对沪深股市2008年1月15号到2012年8月20号这1200对数据进行时间序列分析发现EGARCH模型可以较好的描述沪深股市的波动性特征与杠杆性,然后采用欧式平方距离法选取出t-Copula函数来描述沪深股市的相依性,最后分别用传统的蒙特卡洛模拟方法、t-Copula-GARCH模型、t-Copula-EGARCH模型分别来建模,对求得的VaR进行比较。并对各个结果进行事后检验发现t-Copula-EGARCH在这几种方法求出的VaR最符合实际情况。
[Abstract]:With the development of financial derivatives and information technology, the trading volume of the global financial market is becoming larger and larger. However, the prosperity of the financial market has hidden huge risks, such as the subprime mortgage crisis in 2007 in the United States. In end of 2009, the Greek debt crisis occurred. The financial crisis brought huge losses to investors. Therefore, how to effectively calculate the risk value of financial assets has become a very important problem faced by various financial institutions. Generally, there are three main methods to calculate the risk value of financial assets: sensitivity analysis. Historical income simulation method, Monte Carlo simulation method and so on. The VaR of a single financial asset is easy to calculate, but the calculation of VaR under the joint distribution of multiple financial assets is more difficult. We usually assume that multiple financial assets are dependent on linear correlation and then calculate the value of risk. However, multiple financial assets are not subject to linear correlation. Therefore, this paper introduces Copula theory to describe the correlation of multiple financial assets. From January 15th 2008 to August 20th 2012, we analyze the time series of Shanghai and Shenzhen stock market, and find that EGARCH model can describe the volatility of Shanghai and Shenzhen stock market. Characteristics and leverage. Then using the Euclidean square distance method to select t-Copula function to describe the dependence of Shanghai and Shenzhen stock market. Finally, the traditional Monte Carlo simulation method is used to simulate the t-Copula-GARCH model. The t-Copula-EGARCH model is modeled separately. The VaR obtained is compared and the VaR obtained by t-Copula-EGARCH is found to be the most suitable for the actual situation.
【学位授予单位】：西南财经大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224

【参考文献】