VAR exogenous variable shock

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

lotusfr
Posts: 9
Joined: Mon Jul 13, 2020 1:41 am

VAR exogenous variable shock

Postby lotusfr » Tue Jul 12, 2022 9:23 am

Hi there,

How to create IRF using exogenous variable shocks such as oil, USGDP and China GDP?

Thanks very much

Rong

Capture.PNG
Capture.PNG (11.25 KiB) Viewed 2889 times
Attachments
VAR exgenous variable shock.wf1
(256.95 KiB) Downloaded 89 times

random_user
Posts: 17
Joined: Wed Jun 23, 2021 1:48 pm

Re: VAR exogenous variable shock

Postby random_user » Tue Jul 12, 2022 11:42 am

It's possible to do this, but not easy. It can be done by imposing restrictions on the coefficients in the "VAR Restrictions" tab.
Essentially, you need to 1) add the exogenous variables as endogenous variables, and 2) make sure that, for the equation for the exogenous variable, all the lags on the coefficients of the lags of the other variables are forced to be equal to zero, and 3) make sure you either put the exogenous variables first (if you're using a Choleski identification) or impose a zero restriction of zero in the short-run impulse response matrix, such that the structural shocks over other variables don't affect your exogenous variable.

lotusfr
Posts: 9
Joined: Mon Jul 13, 2020 1:41 am

Re: VAR exogenous variable shock

Postby lotusfr » Wed Jul 13, 2022 1:43 am

random_user wrote:It's possible to do this, but not easy. It can be done by imposing restrictions on the coefficients in the "VAR Restrictions" tab.
Essentially, you need to 1) add the exogenous variables as endogenous variables, and 2) make sure that, for the equation for the exogenous variable, all the lags on the coefficients of the lags of the other variables are forced to be equal to zero, and 3) make sure you either put the exogenous variables first (if you're using a Choleski identification) or impose a zero restriction of zero in the short-run impulse response matrix, such that the structural shocks over other variables don't affect your exogenous variable.


Hi thanks very much.

Could you please provide more details like a screenshot or how to do this using a program?

EViews Matt
EViews Developer
Posts: 560
Joined: Thu Apr 25, 2013 7:48 pm

Re: VAR exogenous variable shock

Postby EViews Matt » Wed Jul 13, 2022 10:42 am

Hello,

Depending on what you're after, it may be easier to use a custom impulse vector that matches the effect of the exogenous variable. For example, the following code simulates a one unit shock to the exogenous variable z.

Code: Select all

wfcreate u 100
series x = nrnd
series y = rnd / 5 + .2
series z = rnd
var v.ls 1 2 x @pch(y) @ c z

matrix v_coefs = v.@coefmat
vector impulse = v_coefs.@row(6)   ' This row holds the estimated coefficents on z.
v.impulse(imp=user, fname=impulse, se=none)


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 23 guests