Empirical Study Of Conditional VaR Based On The Lag And Dull Variable Quantile Regression Model

In most articles, quantile regression model is linear, but in practice, the linear quantile regression model can not suit the practical demand very well. So in this pa-per, the lag and dull variable quantile regression model is presented to estimate the conditional VaR, which is conditioned on the liquidity risk measure. By the empiri-cal analysis, we can find that the lag2and dull variable quantile regression model can better describe real data than the linear quantile regression model and the dull variable quantile regression model based on the liquidity risk measure. And by the backtesting of the conditional value at risk, the conditional value at risk can be better stimulated by the lag2and dull variable quantile regression model. Then we discuss and compare the conditional VaR on different quantiles based on the lag and dull variable quantile regression model. A conclusion is given in the last chapter.