基于分位数回归模型对我国证券市场在险价值度量的实证研究

Based on the quantile regression model, we find that the application of quantile regression model can set the confidence level endogenously in the estimation of parameter of the model, which can overcome the difficulty that the confidence level of can only be determined exogenously. Further positive research on the Shanghai and Shenzhen Stock Exchanges based on quantile regression model find that it can effectively measures the of the Exchanges and depicts the dynamic changes of . We also draw a conclusion that the Shenzhen Stock Exchange exist a big fluctuation than that of Shanghai Stock Exchange and neither of the two markets exists dramatic jump phenomena.