建立var模型时,滞后阶数如何确定啊,aic的值最小还是绝对值最小?做出来的var模型检验发现两个单位根落在单位圆外是不是就不能做脉冲响应函数和方差分解了?望高手解答,这是滞后三阶的结果
LNS LNRE LNMV LNY
LNS(-1) 0.436766 0.372696 0.013089 0.325504
(0.27669) (2.13172) (1.22089) (0.58039)
(1.57857) (0.17483) (0.01072) (0.56084)
LNS(-2) 0.226729 1.737837 3.198096 0.001792
(0.29354) (2.26154) (1.29524) (0.61574)
(0.77241) (0.76843) (2.46912) (0.00291)
LNS(-3) 0.496971 0.834187 -1.271408 0.897694
(0.25072) (1.93163) (1.10629) (0.52591)
(1.98221) (0.43186) (-1.14925) (1.70692)
LNRE(-1) 0.026114 -0.173394 -0.086515 -0.088038
(0.03790) (0.29200) (0.16723) (0.07950)
(0.68903) (-0.59382) (-0.51733) (-1.10739)
LNRE(-2) -0.014390 0.185067 -0.074397 0.051412
(0.03977) (0.30637) (0.17547) (0.08341)
(-0.36188) (0.60406) (-0.42400) (0.61634)
LNRE(-3) -0.045925 -0.321428 -0.018173 -0.079684
(0.04317) (0.33263) (0.19050) (0.09056)
(-1.06374) (-0.96633) (-0.09539) (-0.87988)
LNMV(-1) -0.074770 -0.382381 1.187297 -0.149568
(0.04061) (0.31290) (0.17920) (0.08519)
(-1.84107) (-1.22206) (6.62539) (-1.75569)
LNMV(-2) 0.103847 0.140586 -0.271277 0.090852
(0.06093) (0.46946) (0.26887) (0.12782)
(1.70429) (0.29947) (-1.00896) (0.71080)
LNMV(-3) -0.032854 0.183379 0.019533 0.095162
(0.04601) (0.35445) (0.20300) (0.09650)
(-0.71413) (0.51736) (0.09622) (0.98609)
LNY(-1) -0.106445 0.743270 0.416535 0.184435
(0.13501) (1.04018) (0.59573) (0.28320)
(-0.78843) (0.71456) (0.69920) (0.65125)
LNY(-2) 0.064939 -0.723747 -1.248727 -0.127401
(0.16396) (1.26320) (0.72347) (0.34393)
(0.39607) (-0.57295) (-1.72603) (-0.37043)
LNY(-3) 0.031946 0.693402 -0.738153 -0.092580
(0.14915) (1.14911) (0.65812) (0.31286)
(0.21419) (0.60343) (-1.12160) (-0.29591)
C -0.911106 -19.16131 -2.986014 -2.197932
(1.02974) (7.93361) (4.54377) (2.16004)
(-0.88479) (-2.41521) (-0.65717) (-1.01754)
R-squared 0.994050 0.945178 0.962816 0.964585
Adj. R-squared 0.991406 0.920813 0.946290 0.948846
Sum sq. resids 0.032686 1.940241 0.636424 0.143826
S.E. equation 0.034794 0.268069 0.153529 0.072985
F-statistic 375.9249 38.79218 58.25972 61.28318
Log likelihood 85.43592 3.763808 26.05785 55.80310
Akaike AIC -3.621796 0.461810 -0.652892 -2.140155
Schwarz SC -3.072910 1.010695 -0.104007 -1.591269
Mean dependent 8.329461 7.405163 9.619410 7.151495
S.D. dependent 0.375325 0.952620 0.662464 0.322697
Determinant Residual Covariance 4.02E-10
Log Likelihood 205.6836
Akaike Information Criteria -7.684180
Schwarz Criteria -5.488637