Shock to an exog. variable in a VAR (VARX)

[Fenix]

Hey guys!

I have this problem that I have been struggling with it for over a week. I have a VARX - a vectorautoregression with one exogenous variable (apart from the constant). This variable is stochastic, so I want to see what happens to the system if there is a shock to it but I cannot do it in Eviews. How is that possible there? I've turned the manual upside down and didn't find even a clue regarding VARX. I put the variable as exogenous but after that there is nothing I can do.

I'm on Eviews 5, but I played on the laptop of one of my professors with 7 and didn't see anything changed to the VAR, so I guess they are the same regarding how they treat the exog. variable.

Thanks!

[EViews Gareth]

Not quite sure what you're trying to do, but if you want to see how the endogenous variables change with a shock to an exogenous variable, the only way to do it is by turning your VAR into a Model, solving the model, then change the value of the exogenous variable and go again.

[Fenix]

Thanks Gareth!

Shame to hear that though...
I have a model of this sort - http://econ.la.psu.edu/~hbierens/EasyRe ... s/VARX.PDF This doesn't have to read that as a whole, just a glimps on the first page is sufficient.
Basically the following setup

y_t= c + sum(from i=1 to k_y) A(i) * Y_t-i + sum(from i=0 to k_x) B(i) * X_t-i + e(y)_t

where e(y)_t is white noise
and the X is exogenous to y / such that X is orthogonal to Y and follows its own process

X_t = sum(from i=1 to k_x) D(i) * X_t-i + e(X)_t
e(X)_t is white noise

I want to put a shock to e(X)_t and see what happens and I was hoping there was an *easy* way to do this, since I've never solved before things with Eviews.

[nguyenhuong_ueh]

Dear Fenix,
Now i'm in the same trouble like you. If you founded the solution, please post the answer in this topic. Thanks advance!

[jason83189]

Not quite sure what you're trying to do, but if you want to see how the endogenous variables change with a shock to an exogenous variable, the only way to do it is by turning your VAR into a Model, solving the model, then change the value of the exogenous variable and go again.

Hi Gareth,

Do you know how to add confidence interval after obtaining the response?
I have been trying to find a solution for a long time. Please help.

Thanks.