有同志能回答一两个问题不？

Fitting a VAR or SVAR
var [TS] var Fit vector autoregressive models
svar [TS] var svar Fit structural vector autoregressive models
varbasic [TS] varbasic Fit a simple VAR and graph IRFs or FEVDs
Model diagnostics and inference
varstable [TS] varstable Check the stability condition of VAR or SVAR estimates
varsoc [TS] varsoc Obtain lag-order selection statistics for VARs
and VECMs
varwle [TS] varwle Obtain Wald lag-exclusion statistics after var or
svar
vargranger [TS] vargranger Perform pairwise Granger causality tests after var
or svar
varlmar [TS] varlmar Perform LM test for residual autocorrelation
after var or svar
varnorm [TS] varnorm Test for normally distributed distu

1. 需要是平稳的。但如果y x协整，可以考虑用VECM

如果y x不协整，可以参考Testing for Causality with Wald Tests under Nonregular Conditions. M. Burda 这篇论文。基本思想就是，如果你确定var的滞后阶数为p, 你在var中加入p+1期滞后，然后wald test前p期的联合概率.结果consistent

2. 用varsoc. 观察AIC 或BIC

3. generally, chi2 is used in large sample, while F in finite sample. but they are similar test

[此贴子已经被作者于2008-2-24 22:21:57编辑过]

难道中国没有高手了