Dear Sven, 

Thanks for your timely answer! I guess you understood right. But let me try to be more precise. 

I want to check the effects of a shock within the VAR by restricting the x_1 variable to not move. This is because I want to see if the remaining variables continue to react in the same way. But to this end, I think (not sure) that I have to set the coefficients of x_1 in ALL the other equations equal to 0. In other words, I am trying to figure out the transmission mechanism of the underlying shock, so I have to switch off the x_1 channel. So I want not only the IRF of x_1 to be zero, but I also want that the other IRFs to not consider the reaction of x_1. 

I there a way to do that using the SVAR function?

2013/6/4 Sven Schreiber <>
Am 04.06.2013 15:55, schrieb Gabriela Nodari:
> Dear gretl users,
> I was wondering whether it is possible to restrict the coefficients of
> the estimated structural VAR EX-POST (the estimation).
> In practice, I am trying to do a counterfactual analysis and I want to
> compare IRFs, by restricting one variable to not respond to an
> underlying shock at ALL horizons.
> Suggestions?

If I understand correctly what you mean, if your non-responding variable
is x_1, then you need to throw out all other variables in the x_1
equation. Plus the contemporaneous restriction. Only then will the IRF
be identically zero. That's not ex post, however, but "simply"
restricted estimation.


Gretl-users mailing list