Am 19.10.2010 08:34, schrieb Talha Yalta:
I got this weird output for the Portmanteau test. Maybe worth a
check?
Talha
VAR system, lag order 7
OLS estimates, observations 1989-2008 (T = 20)
Log-likelihood = 189.21287
Determinant of covariance matrix = 2.0779778e-011
AIC = -15.3213
BIC = -13.5290
HQC = -14.9714
Portmanteau test: LB(5) = 76.5391, df = -8
[179769313486231570000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000.0000]
Equation 1: LN_GDP
...
Maybe it's related to 7 lags and 20 obs, which may run already into
d.o.f. problems? Of course the behavior is still a bug.
-sven