论文部分内容阅读
本文基于沪深300股指期货四个不同样本期的1分钟交易数据,比较研究了静态线性的马尔可夫转换自回归模型MSA(Markov Switch Autoregress Model)和动态非线性Symmetrised Joe-Clayton Copula模型在对金融变量间相互关系上建模的适用性。研究结果表明,当期价格波动与滞后一期交易行为间不存在稳定的线性关系,但存在明显的尾部相关结构,并且其相关性在趋势行情中尤为显著。这一结果不但表明,趋势行情中滞后一期的交易行为可以作为当期价格波动的先行指标,还展现出了非线性Copula模型在描述金融变量间的相互关系上的适用性和显著优势。
Based on the 1-minute transaction data in four different sample periods of CSI300 stock index futures, this paper compares the static linear Markov Switch Autoregress Model (MSA) and the dynamic nonlinear Symmetrized Joe-Clayton Copula Applicability of modeling on the relationship between financial variables. The results show that there is no stable linear relationship between the current price volatility and lagged one-time transaction, but there is a significant tail-related structure, and its correlation is particularly significant in the trend market. This result not only shows that lagged one-year trading in the trend market can be used as a prior indicator of the current price volatility, but also shows the applicability and significant advantages of the non-linear Copula model in describing the interrelationships among financial variables.