导　　师： 雷钦礼

学科专业： 020208

授予学位： 硕士

作　　者： ;

机构地区： 暨南大学

摘　　要： 2007年美国次贷危机爆发进而演变为全球金融危机,金融风险监管再次成为焦点,而金融风险监管的关键是对金融风险的度量,VaR方法已经成为金融市场及其他市场的风险度量的主要方法。 VaR和CVaR理论并不是完美无缺的,本文在总结前人的理论的基础上,结合金融风险的特点/(如自回归、波动性、尖峰厚尾/),提出了ARMA-GARCH-Laplace模型来度量我国股市VaR与CVaR值。本文以1997年1月1日至2010年4月15日的上证综指和深圳成分指数各3211个价格指数数据为样本,运用不同分布下的GARCH模型对我国股市风险进行了度量,从返回检验的结果可以看出基于Laplace分布的计量模型优于基于传统的正态分布、T分布及GED分布的计量模型,也进一步证实了我国股市尖峰厚尾的特点。 In 2007 the U.S. subprime mortgage crisis, and then evolved into the global financial crisis, financial risk control Once again become the focus of supervision, but financial risks and the key lies in the financial risk based. VaR/(value-at-risk/)method has become the mainstream for risk management and regulatory in the financial market. However, VaR and CVaR theory is not perfect, this article summarizes the basis of previous theory, combining the characteristics of financial risk /(such as regression, volatility, a fat tail/), put forward this article ARMA-GARCH-Laplace Model to measure the value of Chinese stock market VaR and CVaR. In this paper,1 January 1997 to April 15,2010 the Shanghai Composite Index and Shenzhen Component Index of the 3211 price index data for the sample,susing different distribution of the GARCH model to a measure of stock market risk, from the back testing results that the measurement model based on the Laplace distribution is better than the econometric models based on the traditional normal distribution, T distribution and the GED distribution, also confirmed China's stock market peak and fat tail features.

关 键 词： 风险 在险价值 条件在险价值 分布

分 类 号： [F224 F831.59]

领　　域： [经济管理] [经济管理]