基于非对称结构与长记忆的股票市场风险测度研究

Study on Dynamic Risk Management in Financial Markets Based on Asymmetry Structure and Long Memory Characters

针对金融市场条件收益存在的"有偏胖尾"分布与非对称波动性特征以及长记忆特征等典型事实特征,运用ARFIMA-FIGARCH-SKST模型等来测度股市动态风险,并通过规范的返回测试检验中的LRT和DQR方法实证考察了测度模型的可靠性。得到了一些非常有价值的实证结果:有无长记忆约束的非对称结构风险模型在中国大陆沪深股市动态风险测度能力上并无实质性差异;ARFIMA-FIAPARCH-SKST模型能够准确测度股市的动态风险;股票市场极端风险的测度尤其不能放弃非对称结构的这一约束条件。

To address skew-fated tail distribution,asymmetry volatility structure,long memory character of financial return series,we apply ARFIMA-FIGARCH-SKST models to measure dynamic risk for financial markets,and test the accuracy of risk measuring models by Likelihood Ratio Test（LRT） and Dynamic Quantile Regression（DQR） method.Some important results are concluded： different risk models show difference for different markets;asymmetry risk model with long memory character don＇t show essential difference from that without long memory character in Chinese stock markets;ARFIMA-FIAPARCH-SKST model can measure dynamic risks of different markets;asymmetry structure,as a constraint,should not be abandoned in measuring extreme risks of financial markets.