历史分解要在excel或Eviews中自己算,几句话也说不清楚!看看文献就会了!也就是看看别人的文章!
你问的两个问题要根据数据自己尝试!A或B矩阵自己根据数据和建立的模型设定,不需要编程!那些约束要根据你的数据、检验结果和模型决定,别弄错!不然后面的结果就全错了!
也就是常说中的论文中的重大错误。
另外,建议你首先把VAR和SVAR的思想和数学模型看懂,你什么就都会了!
Structural VAR Estimates
Date: 04/29/09 Time: 13:55
Sample (adjusted): 1995Q3 2007Q4
Included observations: 50 after adjustments
Estimation method: method of scoring (analytic derivatives)
Convergence achieved after 10 iterations
Structural VAR is over-identified (3 degrees of freedom)
Model: Ae = Bu where E[uu']=I
Restriction Type: short-run text form
@e1=1.71*@e3+@u1
@e2=+c(1)*@e1+@u2
@e3=+c(2)*@e1+c(3)*@e2+@u3
where
@e1 represents LNTAX residuals
@e2 represents LNG residuals
@e3 represents LNGDP residuals
WARNING: B matrix is fixed (structural innovation variances not estimated)!!!
Coefficient Std. Error z-Statistic Prob.
C(1) -6.668522 0.668411 -9.976680 0.0000
C(2) 0.138208 0.944290 0.146362 0.8836
C(3) 27.30830 2.527545 10.80428 0.0000
Log likelihood 97.74618
LR test for over-identification:
Chi-square(3) 147.1573 Probability 0.0000
Estimated A matrix:
1.000000 0.000000 -1.710000
6.668522 1.000000 0.000000
-0.138208 -27.30830 1.000000
Estimated B matrix:
1.000000 0.000000 0.000000
0.000000 1.000000 0.000000
0.000000 0.000000 1.000000


雷达卡


京公网安备 11010802022788号







