基于价格极差-ARMA-GARCH-POT模型的风险价值研究

发布时间：2018-01-18 00:13

本文关键词：基于价格极差-ARMA-GARCH-POT模型的风险价值研究　出处：《西华师范大学》2016年硕士论文　论文类型：学位论文

【摘要】：近年来,随着信息科技迅猛发展,中国证券市场也在发生着日新月异的变化。然而我国的金融市场是一个较新兴的市场,存在着市场结构不健全、投机性较强等问题。除此之外,全球一体化进程不断的深入,各种政策和消息在给全球金融市场带来冲击时,也给我国的金融市场带来冲击,证券市场就会随之受到影响。随着我国证券市场的不断发展,面临的风险也越来越多。如何准确度量风险,将损失降到最低就变得越来越重要。而当今世界度量金融资产风险价值的主流方法便是VaR模型,大量的国内外学者都在研究如何用Va R模型来度量风险,提高风险价值的有效性,这对度量金融风险具有重要的意义。因此本文在前人的基础上,将引入价格极差的GARCH模型与方差-协方差法和极值理论三者相结合改进VaR模型,旨在提高VaR值的有效性。本文从金融市场风险的背景展开研究,详细阐述了风险价值VaR的定义和基本的计算方法,介绍了应用最广泛的估计VaR的组合模型,即GARCH模型和方差-协方差模型的组合,从这个模型中可以知道估计VaR需要知道两个量:即对金融时间序列建立GARCH模型后得到的标准差序列和时间序列概率分布的分位数。因此可以从这两方面入手改进模型,运用包含股票收盘价、最高价和最低价的极差-GARCH模型来提高标准差的有效性;在运用方差-协方差与GARCH模型构成的组合模型估计VaR时,通常都是在正态假设的情况下进行的,但是研究表明时间序列往往具有尖峰厚尾的特点,所以基于正态分布假设求VaR的方法一般就会低估尾部风险,而极值理论是直接处理时间序列数据的尾部,并且不需要对损失数据预先假设服从任何分布,因此由极值理论求出的分位数更为有效。将这两个模型组合在一起构成了新模型:价格极差---POTGARCHARMA模型,最后经过实证分析,价格极差---POTGARCHARMA模型求出的VaR值,比未改进前的模型估计出的VaR值的失败率更接近于显著性水平,表明了新模型价格极差---POTGARCHARMA模型确实提高了VaR值的有效性,说明文中对VaR模型的改进是可行的。本文最后对实证分析得出的结论进行总结,并详细阐述了本文的不足之处,在不足之处的基础上,提出了以后的一些研究方向。
[Abstract]:In recent years, with the rapid development of information technology, China's securities market is also changing with each passing day. However, China's financial market is a relatively new market, there is a market structure is not perfect. In addition, the process of global integration continues to deepen, various policies and news to the global financial market impact, but also to our financial market impact. The securities market will be affected. With the continuous development of China's securities market, there are more and more risks. How to accurately measure the risk. It becomes more and more important to minimize losses, and the mainstream method of measuring the risk value of financial assets in the world today is the VaR model. A large number of scholars at home and abroad are studying how to use the VaR model to measure risk and improve the effectiveness of the value of risk, which is of great significance to the measurement of financial risk. The GARCH model of price range is combined with the variance-covariance method and the extreme value theory to improve the VaR model. In order to improve the effectiveness of VaR, this paper studies the background of risk in financial markets, and expounds the definition and basic calculation method of VaR in detail. In this paper, the most widely used combinatorial model for estimating VaR is introduced, that is, the combination of GARCH model and variance-covariance model. From this model, you can see that estimating VaR needs to know two quantities:. That is the quantiles of the standard deviation series and the probability distribution of the time series after the establishment of the GARCH model for the financial time series, so we can improve the model from these two aspects. The GARCH model, which includes the closing price, the highest price and the lowest price, is used to improve the effectiveness of the standard deviation. When using the combination of variance-covariance and GARCH model to estimate VaR, it is usually carried out under the normal assumption, but the research shows that the time series often have the characteristics of peak and thick tail. Therefore, the method of calculating VaR based on normal distribution hypothesis generally underestimates tail risk, while extreme value theory deals with the tail of time series data directly, and does not presuppose any distribution of lost data. Therefore, the quantiles derived from the extreme value theory are more effective. The two models are combined to form a new model: price difference-POTGARCHARMA model, and finally through empirical analysis. The VaR value calculated by POTGARCHARMA model is closer to the significant level than the VaR value estimated by the unimproved model. It shows that the price of the new model is very poor-POTGARCHARMA model does improve the validity of VaR value. The improvement of VaR model is feasible. Finally, the conclusion of empirical analysis is summarized, and the inadequacies of this paper are described in detail, on the basis of deficiency. Some future research directions are proposed.
【学位授予单位】：西华师范大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F224

【参考文献】