沪深300股指期现货市场间相依度测度实证研究

考虑金融市场间非线性、非正态性假设下的交叉相关性特征,基于复杂性理论视角,引入更符合金融市场实际的交叉相关性统计量检验、去趋势交叉相关性分析（DCCA）及多重分形去趋势交叉相关性分析（MF-DCCA）等方法对沪深300股指期现货市场间的相依度测度进行定量研究.研究结果表明：我国沪深300股指期现货市场不但自身具有长记忆性及多重分形特征,且两市场间也存在着显著的交叉相关性和交叉多重分形特征.上述研究结果将为深入研究多市场间非线性依赖关系和复杂特性机理提供有益参考....

Full description

Saved in:
Bibliographic Details
Published in东北大学学报（自然科学版） Vol. 38; no. 1; pp. 148 - 152
Main Author 苑莹 王梦迪 张同辉 樊晓倩
Format Journal Article
LanguageChinese
Published 东北大学工商管理学院,辽宁 沈阳,110169%华夏银行 沈阳分行,辽宁 沈阳,110000 2017
Subjects
去趋势交叉相关性分析
多重分形
滑动时间窗
相依度
长程相关性
多重分形
detrended cross-correlation analysis
multifractal
dependency
sliding timewindow
long-term correlation
相依度
去趋势交叉相关性分析
滑动时间窗
长程相关性
Online AccessGet full text
ISSN1005-3026
DOI10.3969/j.issn.1005-3026.2017.01.030

Cover

More Information
Summary:考虑金融市场间非线性、非正态性假设下的交叉相关性特征,基于复杂性理论视角,引入更符合金融市场实际的交叉相关性统计量检验、去趋势交叉相关性分析（DCCA）及多重分形去趋势交叉相关性分析（MF-DCCA）等方法对沪深300股指期现货市场间的相依度测度进行定量研究.研究结果表明：我国沪深300股指期现货市场不但自身具有长记忆性及多重分形特征,且两市场间也存在着显著的交叉相关性和交叉多重分形特征.上述研究结果将为深入研究多市场间非线性依赖关系和复杂特性机理提供有益参考.
Bibliography:YUAN Ying1 ,WANG Meng-di2, ZHANG Tong-hui1 ,FAN Xiao-qian1（1. School of Business Administration,Northeastern University,Shenyang 110169, China; 2. Shenyang Branch, Huaxia Bank,Shenyang 110000, China.）
21-1344/T
Considering the financial markets＇ characteristics of the cross-correlation in the case of nonlinearity and non-normality, and from the perspective of complexity theory, the crosscorrelation statistic test,detrended cross-correlation analysis（ DCCA） and multifractal detrended cross-correlation analysis（ MF-DCCA） are mainly applied to quantitatively study the dependency of the CSI 300 index futures market and spot market,which are more in line with the financial market. The results showthat long-term memory and multifractal characteristics exist in the CSI300 index futures and spot markets respectively. Meanwhile,there are obvious cross-correlation characteristics between futures and spot markets,and the cross-correlation is multifractal. The results can provide a better reference for the nonlinear depend
ISSN:1005-3026
DOI:10.3969/j.issn.1005-3026.2017.01.030