Page 1 of 1

confidence interval for impulse response

Posted: Mon Jun 19, 2017 3:23 am
by e.kheirandish
Hello

I estimated a VAR reduced form via OLS and generated impulse responses by programming. But I couldn't calculate confidence interval for impulse responses :( Please guide me how can I compute it. Please attention, Because of I couldn't use Eviews package for estimation of VAR, now I have only coefficient matrix and OLS estimation results, impulse response matrix and residual variance covariance matrix.

Many thanks in advance

Re: confidence interval for impulse response

Posted: Mon Jun 26, 2017 8:59 am
by EViews Matt
Hello,

Is the issue that the standard error bounds weren't generated when you used the .impulse procedure?

Re: confidence interval for impulse response

Posted: Tue Jun 27, 2017 8:41 pm
by e.kheirandish
Hello,

Is the issue that the standard error bounds weren't generated when you used the .impulse procedure?

No, since my model was an unusual model, I couldn't use the .impulse procedure. So, I generated responses manually (via Programming) and now I can't calculate standard error bounds for responses. please guide me!

very thanks in advance

Re: confidence interval for impulse response

Posted: Thu Jun 29, 2017 11:59 am
by EViews Matt
There are several methods for generating the bounds. You might take a look at a (relatively) simple overview, such as http://pareto.uab.es/lgambetti/VAR_Forecasting.pdf, to see what methodology you'd prefer.

Re: confidence interval for impulse response

Posted: Sat Jul 15, 2017 3:05 am
by e.kheirandish
Thanks a lot for your attention