基于GARCH-EVT-Copula模型对股指对冲交易的波动性研究
发布时间：2018-07-24 08:15
【摘要】：由于经济结构转型,金融市场波动日趋激烈,想要获得高额收益,意味着承担更大的风险。组合投资规避的是非系统性风险,股指期货可以规避系统性风险。对冲交易最大限度的控制了股票市场的风险。现在的市场逐步迈入对冲交易时代,但由于期货市场套期保值企业较少,更多的是投机散户,导致期货市场的发展受限。再加上交易系统数据支持不够完善,且国内量化基金无法进行瞬间利差交易,单纯的对冲交易存在着较大的风险性。市场结构相对不稳定,受股票市场和期货市场各自波动的影响也给对冲交易带来了 一定的风险性。因此建立有效的金融模型,对对冲交易的波动性研究尤其重要。文章把对冲交易的股指和期指做为组合分析来研究。通过对收益序列的统计检验分析,应用GARCH-EVT-Copula模型研究对冲交易的波动性。文章不仅对模型的定义给予清晰的介绍,而且针对模型的特性,对模型的实际应用给予了侧重研究和部分改进。文章通过对波动性定量建模,将有效解释单一资产分布的条件异方差模型和注重极端风险的EVT理论结合构建股指和期指的边缘分布。先通过GARCH过程得到标准化残差序列。然后应用EVT理论,对残差序列上下尾部进行拟合,中间部分创新选用t-分布进行拟合。再应用灵活的二元函数连接边缘分布,构造联合分布。最后,通过t-Copula过程的参数生成模拟收益率序列,估计了对冲风险VaR值。通过返回检验,对模型的有效性进行了分析。通过对数据的分析和统计检验,应用改进的GARCH+EVT+tCopula模型估计对冲交易的在值风险VaR。估计得到在显著水平0.1和0.05的情况下的VaR值。并做了返回检验和相关的模型检验,得出GARCH+EVT+tCopula模型能有效评估对冲交易的风险。此外,文章还选用其他三种相关模型来在相同的显著水平下,对VaR进行估计,用来对比改进处理后的GARCH+EVT+tCopula模型的拟合效果。结果证实了,通过EVT拟合尾部后EVT更好的把控了极端风险。GARCH-EVT-Copula模型拟合的效果更接近于实际,为对冲交易分析提供了较为有效的模型。
[Abstract]:Due to the structural transformation of the economy, financial market volatility is becoming increasingly fierce, to achieve high returns, means to take on greater risk. Portfolio to avoid non-systemic risk, stock index futures can avoid systemic risk. Hedge trading maximizes the risk of the stock market. Now the market has gradually entered the era of hedging, but because of the futures market hedging enterprises are fewer, more speculative retail investors, resulting in the development of the futures market is limited. In addition, the trading system data support is not perfect, and the domestic quantitative funds can not carry out the instantaneous interest rate difference trading, the pure hedge trading has a greater risk. The market structure is relatively unstable, which is influenced by the fluctuation of stock market and futures market. Therefore, it is very important to establish an effective financial model to study the volatility of hedging transactions. This paper studies the index and index of hedge trading as a combination analysis. Based on the statistical analysis of return series, the volatility of hedge trades is studied by using GARCH-EVT-Copula model. This paper not only gives a clear introduction to the definition of the model, but also focuses on the research and partial improvement of the practical application of the model according to the characteristics of the model. Based on the quantitative modeling of volatility, the conditional heteroscedasticity model which effectively explains the distribution of single assets and the EVT theory focusing on extreme risk are combined to construct the marginal distribution of stock index and futures index. The standardized residuals are obtained by GARCH process. Then the EVT theory is used to fit the upper and lower tail of the residual sequence, and the t- distribution is used to fit the middle part. Then a flexible binary function is used to connect the edge distribution to construct the joint distribution. Finally, the VaR value of hedging risk is estimated by generating the simulated return sequence by the parameters of the t-Copula process. Through the return test, the validity of the model is analyzed. Based on the data analysis and statistical test, the improved GARCH EVT tCopula model is applied to estimate the VaR of hedging transactions. The VaR values were estimated to be 0. 1 and 0. 05 at the significant level of 0. 1 and 0. 05 respectively. A return test and a related model test are made, and it is concluded that the GARCH EVT tCopula model can effectively evaluate the risk of hedging transactions. In addition, the other three models are used to estimate the VaR at the same significant level, and to compare the fitting effect of the improved GARCH EVT tCopula model. The results show that after EVT fitting, EVT can better control the extreme risk. GARCH-EVT-Copula model is closer to reality, which provides a more effective model for hedge trade analysis.
【学位授予单位】：天津财经大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F224;F724.5
[Abstract]:Due to the structural transformation of the economy, financial market volatility is becoming increasingly fierce, to achieve high returns, means to take on greater risk. Portfolio to avoid non-systemic risk, stock index futures can avoid systemic risk. Hedge trading maximizes the risk of the stock market. Now the market has gradually entered the era of hedging, but because of the futures market hedging enterprises are fewer, more speculative retail investors, resulting in the development of the futures market is limited. In addition, the trading system data support is not perfect, and the domestic quantitative funds can not carry out the instantaneous interest rate difference trading, the pure hedge trading has a greater risk. The market structure is relatively unstable, which is influenced by the fluctuation of stock market and futures market. Therefore, it is very important to establish an effective financial model to study the volatility of hedging transactions. This paper studies the index and index of hedge trading as a combination analysis. Based on the statistical analysis of return series, the volatility of hedge trades is studied by using GARCH-EVT-Copula model. This paper not only gives a clear introduction to the definition of the model, but also focuses on the research and partial improvement of the practical application of the model according to the characteristics of the model. Based on the quantitative modeling of volatility, the conditional heteroscedasticity model which effectively explains the distribution of single assets and the EVT theory focusing on extreme risk are combined to construct the marginal distribution of stock index and futures index. The standardized residuals are obtained by GARCH process. Then the EVT theory is used to fit the upper and lower tail of the residual sequence, and the t- distribution is used to fit the middle part. Then a flexible binary function is used to connect the edge distribution to construct the joint distribution. Finally, the VaR value of hedging risk is estimated by generating the simulated return sequence by the parameters of the t-Copula process. Through the return test, the validity of the model is analyzed. Based on the data analysis and statistical test, the improved GARCH EVT tCopula model is applied to estimate the VaR of hedging transactions. The VaR values were estimated to be 0. 1 and 0. 05 at the significant level of 0. 1 and 0. 05 respectively. A return test and a related model test are made, and it is concluded that the GARCH EVT tCopula model can effectively evaluate the risk of hedging transactions. In addition, the other three models are used to estimate the VaR at the same significant level, and to compare the fitting effect of the improved GARCH EVT tCopula model. The results show that after EVT fitting, EVT can better control the extreme risk. GARCH-EVT-Copula model is closer to reality, which provides a more effective model for hedge trade analysis.
【学位授予单位】：天津财经大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F224;F724.5
【相似文献】
相关期刊论文 前10条
1 韩明;Copula——一个新的计量经济工具[J];统计与信息论坛;2004年05期
2 李健伦,方兆本,鲁炜,李红星;Copula方法与相依违约研究[J];运筹与管理;2005年03期
3 单国莉,陈东峰;一种确定最优Copula的方法及应用[J];山东大学学报(理学版);2005年04期
4 罗俊鹏;;基于Copula的金融市场的相关结构分析[J];统计与决策;2006年16期
5 李霞;曾霞;侯兵;;copula的构造以及copula之间关系的研究[J];商丘师范学院学报;2006年05期
6 孙志宾;;混合Copula模型在中国股市的应用[J];数学的实践与认识;2007年20期
7 李娟;戴洪德;刘全辉;;几种Copula函数在沪深股市相关性建模中的应用[J];数学的实践与认识;2007年24期
8 许建国;杜子平;;非参数Bernstein Copula理论及其相关性研究[J];工业技术经济;2009年04期
9 杜子平;闫鹏;张勇;;基于“藤”结构的高维动态Copula的构建[J];数学的实践与认识;2009年10期
10 王s，
本文编号：2140775
本文链接：https://www.wllwen.com/jingjilunwen/jiliangjingjilunwen/2140775.html
