>A criticism VARs face is that they are atheoretical. They are not built on some economic theory that imposes a theoretical structure to the equations. Every variable is assumed to influence every other variable in the system, which makes direct interpretation of the estimated coefficients very difficult.<p>This is maybe me needing a refresher, but in what way does this make interpretation of the coefficients difficult?<p>I'd much rather side with Sims on this:
Sims advocated VAR models as providing a theory-free method to estimate economic relationships, thus being an alternative to the "incredible identification restrictions" in structural models. [1]<p>I was also expecting a bit more material on Granger causality, as I remember using that as the final nail in testing predictive validity.<p>[1] <a href="https://en.wikipedia.org/wiki/Vector_autoregression#Applications" rel="nofollow">https://en.wikipedia.org/wiki/Vector_autoregression#Applicat...</a>