价格持续期的非对称对数ACD模型及其应用

Asymmetric Logarithmic ACD Model for Price Duration and Its Applications

通过建立两状态价格持续期的非对称对数ACD模型,刻画价格持续期过程对不同价格状态的不对称依赖关系和对未预期到的短价格持续期冲击与未预期到的长价格持续期冲击的不对称响应过程,并在该模型中引入买卖价差和交易规模变量,检验市场微观结构理论相关假说。模型拟合结果表明,价格上升和价格下降两种状态对价格持续期的影响不同,未预期到的短价格持续期冲击对价格持续期有正面影响,而未预期到的长价格持续期冲击对价格持续期有负面影响;对市场微观结构信息模型的实证分析表明,滞后买卖价差和滞后交易量具有信息含量,它们与价格持续期显著负相关;隐藏交易假说没有得到实证的支持,在选取样本中大规模交易量比中等规模交易量对价格持续期有更加显著的影响。

Developing two-state asymmetric logarithmic ACD model for price duration to describe the variation of conditional expected price duration asymmetrical depending on price state,and the asymmetric response processes following small and large innovations shock,then incorporate microstructure variables including bid-ask spread,trading volume and size into the Asymmetric ACD model in order to test related hypothesis in market microstructure literature. The modeling result shows the distinct impact of price increase and decrease state on price duration,small innovations have positive effect on conditional duration where large negative. The empirical analysis of related market microstructure hypotheses shows the lagged bid-ask spread and trading volume have information content; they＇re significantly negatively correlated to the conditional price duration. Furthermore,it＇s the large trading size not medium size have more significant impact on price duration; and there is no convincing evidence in favor of the stealth trading hypotheses according to the empirical analysis.