Can I interpret the coefficients in a VAR model in the same way as I do in a normal OLS regression?
$\begingroup$
$\endgroup$
-
$\begingroup$ answer by a prof from my department (University of Leicester): "Yes you can interpret the coefficients in exactly the same way" $\endgroup$ – Johannes Jul 30 '12 at 13:06
-
1$\begingroup$ Probably more informative to look at the impulse response function of the VAR. $\endgroup$ – Jase Jan 11 '13 at 12:01
$\begingroup$
$\endgroup$
It's often pretty hard interpret the coefficients of a VAR, specially if it includes many variables and lags. As one lag of a variable says one thing and another the opposite, there are no clear dynamics between the variables you wish to investigate, usually a VAR is accompanied with tools like the impulse response function and forecast error variance decomposition.
