VAR模型的应用

Р[-0.63427]Р[-1.01636]РCР 0.147718Р 0.577199Р (0.17390)Р (0.17902)Р[ 0.84946]Р[ 3.22423]Р R-squaredР 0.998740Р 0.998174Р Adj. R-squaredР 0.998530Р 0.997870Р Sum sq. residsР 0.054303Р 0.057549Р S.E. equationР 0.047567Р 0.048968Р F-statisticР 4755.470Р 3280.140Р Log likelihoodР 49.91775Р 49.07589Р Akaike AICР-3.097776Р-3.039717Р Schwarz SCР-2.862035Р-2.803976Р Mean dependentР 7.591198Р 7.440455Р S.D. dependentР 1.240583Р 1.060976Р Determinant resid covariance (dof adj.)Р 3.32E-06Р Determinant resid covarianceР 2.27E-06Р Log likelihoodР 106.1169Р Akaike information criterionР-6.628749Р Schwarz criterionР-6.157267Р 从表中可以看出VAR模型的参数估计大多显著。Р3.4 VAR模型的检验РVAR模型的检验包括VAR模型的平稳性检验,以及残差的独立性检验。Р 图5 平稳性检验结果Р因此VAR模型不满足平稳性的条件Р所以残差不存在自相关性,满足独立性假设。Р3.5 VAR模型的预测Р下面是动态预测结果Р4.VAR模型的应用Р4.1 格兰杰因果检验Р本例中VAR模型的滞后阶数为2