市场流动性与市场预期的动态相关结构研究——基于ARMA-GJR-GARCH-Copula模型分析

本文在兼顾"时间尺度"和"价格尺度"双重因素下构建了标准化的市场流动性测度,并利用时变信息熵方法提出了一类市场预期的新指标。将ARMA-GJR-GARCH模型与时变Copula模型相结合分析了市场流动性与市场预期之间的动态相关结构。利用2009年1月~2014年9月中国股市日度数据进行实证分析,结果表明:市场流动性和市场预期存在较明显的持续性和负向"杠杆效应",通过LL、AIC和BIC三种准则比较发现时变正态Copula模型的拟合效果最好,时变相关性分析说明市场流动性和市场预期长期内保持着负相关的总体态势,欧债危机期间时变相关系数在正负状态间转换频繁,其相关结构出现了几次较大的变点,但正常时期两者之间的相关程度并不显著。该结论对于监管部门在危机期间及时引导市场预期,增强市场流动性从而减少危机传染和缓释金融风险非常重要。

Dynamic Correlation Structure Between Market Liquidity and Market Expectation:An Analysis Based on ARMA-GJR-GARCH-Copula Model

A standardized market liquidity measure is construeted by considering "time scale" and "price scale" at the same time, and a new index of market expectation is given by use of time-varying information entropy method. The dynamic correlation structure between market liquidity and market expectation could be analyzed by combining ARMA-GJR-GARCH model with time-varying Copula model. The Chinese stock daily data from January 2009 to September 2014 are utilized for empirical analysis. The results indicate that there exist some significant persistent and negative "leverage effect" for market liquidity and market expectation, the time-varying normal Copula model is optimal through comparative analysis under the LL、AIC and BIC principles, and time-varying relevant analysis shows that in long term, market liquidity and market expectation have kept overall trend of negative relevant, and the time-varying correlation coefficient has frequently switched between positive and negative position,and there exist some big change-points in correlation structure during the Europe-US sovereign debt crisis, while little relevant relationship in normal period. All of these are of great importance for regulators to timely guide market expectation and enhance market liquidity during the crisis, which may reduce the crisis contagion and release the financial risk.