摘要
选择股票市场2014年1月8日到2016年8月2日上证指数(SZCZ)收盘价作为样本进行研究,建立了残差序列服从不同分布下的ARMA(2,2)-GARCH(1,1)类模型,并进行实证分析及预测,通过比较发现ARMA(2,2)-GARCH(1,1)类模型能够有效处理股票数据的波动集聚性。在此基础上,计算出不同置信水平下ARMA(2,2)-GARCH(1,1)类模型的风险测度值(VaR),与历史模拟法、改进的历史模拟法下的VaR值进行比较,得出ARMA(2,2)-GARCH(1,1)类模型更有效的反应市场风险。
In this paper, the author chooses the closing price of the Shanghai index ( SZCZ) from January 8, 2014 to August 2, 2016 as samples. First of all, the author establishes the ARMA (2,2) -GARCH (1,1) mod-el of the different distribution of the residual sequence,and the empirical analysis and prediction. By compari-son ,the author found that the ARMA (2,2) -GARCH (1,1) model can effectively deal with the volatility clus-tering of stock data. Secondly, the author calculates the risk value (VaR) of ARMA (2,2) -GARCH (1,1 ) class model under different confidence levels, and compares with VaR value in historical simulation and improved historical simulation method. Finally, the author concludes that the ARMA (2,2) -GARCH (1,1 ) model is more effective in response to market risk.
出处
《大庆师范学院学报》
2017年第5期23-27,共5页
Journal of Daqing Normal University
基金
安徽省教育厅自然科学研究项目(KJHS2016B04
KJHS2017B09)
黄山学院自然科学研究项目(2015xkj005)