Scaling Impulse Respone

[alirezan321]

Hi there,

I have a VAR model for monetary policy, I want to get the impulse response of output to exact 1 percent shock in federal fund rate. How do I do this? The normal VAR in eview, scales down the response to the standard deviation of the residuals.

Ali.

[alirezan321]

Any reply??

[startz]

Any reply??

Have you looked at the Impulse Definition tab?

[rpn4]

I share his question... In fact, Eviews offers the option of a unit shock to the residuals, but not factorized. General Impulse Responses, or Choleski decompositions, are only available for a standard error shock, not for a unit shock. Any suggestion on how to tackle this issue?

[EViews Glenn]

Any desired shocks not supported by the standard EViews settings may be obtained using a user-specifed vector or matrix.

[Houda Boubaker]

Hey,
I have the same problem as Ali, and i want to ask him if he found a solution to scale the impulse responses. In fact EViews give a standard scale for the reponses and there is no option on it to change the scale into percentage!!!! Please I need an answer.
Thanks or all.
Houda.

[nadja123]

Hey folks, if I remember well, you need to estimate your model in log differences to obtain the IRFs in percentages. But don't take this for granted, check it out first. And btw, don't worry, you'll never obtain help on such issues from the forum moderators. I myself found the VAR documentation and features quite :( ...

[Houda Boubaker]

Hey Nadja123,
Thanks for the reply, but should we transform the variables in log difference?, means (we transform in logaritm form than we make the differenciation following the unit root stationarity test).
**************************************
Houda BEN HADJ BOUBAKER
Phd Student
ISG Tunis
BESTMOD Laboratory
E-mail: Houda.benhadjboubaker@isg.rnu.tn
boubakerhouda@gmail.com
boubakerhouda2004@yahoo.fr
Phone: (+216) 22 630 818
**************************************