Page 1 of 1

Bayesian VAR credibility interval

Posted: Tue Jan 06, 2015 4:13 pm
by amouna82
Hi everyone,

I would appreciate if you can explain to me how to compute the credibility interval (confidence interval) of a bayesain VAR. I am using Eviews8 and it give only the impulse responses without the credibility interval.

Thank you in advance,

Imen

Re: Bayesian VAR credibility interval

Posted: Tue Jan 06, 2015 4:23 pm
by EViews Gareth
EViews does not include CIs for BVARs.

Re: Bayesian VAR credibility interval

Posted: Tue Jan 06, 2015 4:24 pm
by amouna82
Thank you for the reply. Is there any way to compute them manually?

Re: Bayesian VAR credibility interval

Posted: Fri Jan 09, 2015 2:23 pm
by amouna82
Could you please answer to my question? Thank you

Re: Bayesian VAR credibility interval

Posted: Fri Jan 09, 2015 2:37 pm
by EViews Gareth
I don't know.

Re: Bayesian VAR credibility interval

Posted: Fri Jan 09, 2015 3:02 pm
by startz
While I haven't read the BVAR documentation carefully, I think that for at least some of the priors the posterior is normal. In this case the credible intervals are the usual mean plus or minus 1.96 standard errors.

Re: Bayesian VAR credibility interval

Posted: Sat Jan 10, 2015 2:08 am
by amouna82
Yes. It is true, the posterior distribution is normal because I am using Minessota priors. But How can I extract the standard errors?

Re: Bayesian VAR credibility interval

Posted: Sat Jan 10, 2015 8:42 am
by startz
Aren't the standard errors in parentheses as usual?