【摘要】 为了刻画流动性对金融市场风险的影响，并且考虑到金融数据的尖峰厚尾性，对传统的门限UHF-GARCH模型进行扩展，加入持续期门限和对数持续期门限效应，基于标准T分布建立门限UHF-GARCH-MD模型和门限UHF-GARCH-LD模型，结合交易的及时性和交易深度刻画金融市场流动性风险，评价并比较两类门限模型的拟合优度和流动性风险度量精度.实证发现，持续期杠杆效应对金融市场流动性风险存在显著影响，门限UHFGARCH-LD模型不仅刻画了金融数据的尖峰厚尾性，并且捕捉高速流动性和低速流动性对金融市场风险的不对称影响，可有效地度量流动性风险，为金融市场风险研究提供了流动性新视角，也可为市场风险防控提供参考.更多还原

【Abstract】 In order to portray the impact of liquidity on financial market risk,and considering the leptokurtosis characteristics of financial data,the traditional threshold UHF-GARCH model is extended by adding duration threshold and logarithmic duration threshold effects. The threshold UHF-GARCH-MD model and the threshold UHF-GARCH-LD model are established based on the standard T distribution. Combining the timeliness of transactions and the depth of transactions to portray the liquidity risk of financial markets,the goodness of fit and the accuracy of liquidity risk measurement of the two types of threshold models are evaluated and compared. The empirical results show that the duration leverage effect has a significant impact on the financial market liquidity risk. The threshold UHF-GARCHLD model not only portrays the leptokurtosis characteristics of financial data,but also captures the asymmetric effects of high-speed liquidity and low-speed liquidity on financial market risk. In addition,the model can effectively measure liquidity risk,provide a new perspective of liquidity for financial market risk research,and provide a reference for market risk prevention and control.更多还原