analytic error bounds for IRFs

For requesting general information about EViews, sharing your own tips and tricks, and information on EViews training or guides.

Moderators: EViews Gareth, EViews Moderator

nadja123
Posts: 72
Joined: Thu Aug 06, 2009 10:43 am

analytic error bounds for IRFs

Postby nadja123 » Wed Aug 26, 2009 4:09 am

Hi,
please can somebody tell me how does EViews compute analytic (asymptotic) standard errors to VAR impulse responses? I would appreciate the exact formula, or eventually concrete reference to the formula in some EViews documentation.
I have posted this question earlier, under Estimation, however, without response. I bought full version of EViews 6 on own account, so I would find nice if I knew answer to this question.
Thanks!
nadja123

EViews Glenn
EViews Developer
Posts: 2671
Joined: Wed Oct 15, 2008 9:17 am

Re: analytic error bounds for IRFs

Postby EViews Glenn » Wed Aug 26, 2009 12:15 pm

It's a pretty standard calculation. See the Lutkepohl reference that we cite.

nadja123
Posts: 72
Joined: Thu Aug 06, 2009 10:43 am

Re: analytic error bounds for IRFs

Postby nadja123 » Wed Aug 26, 2009 1:48 pm

Better no comment on EViews helpfile, particularly the chapter on VARs, a complete... mess. I had time to recalculate VAR as such and IRFs in Matlab to figure out what EViews does, but do not really have time to recalculate by trial and error everything. Therefore I asked. I imagined a sort of bunch of formulas with a commentary could be somewhere in EViews documentation. Okay, indeed, after purchasing a soft I'm going to purchase a book or restricted article to find out what that software precisely computes. Thank you.

EViews Glenn
EViews Developer
Posts: 2671
Joined: Wed Oct 15, 2008 9:17 am

Re: analytic error bounds for IRFs

Postby EViews Glenn » Wed Aug 26, 2009 2:07 pm

I am sorry that you are disappointed. While there are many places where the EViews documentation could be improved, I hope that you understand that it is impossible for us to provide complete formulae for every statistical calculation in EViews. This is especially true where there are no special complications over the descriptions in the literature. Just as textbooks cannot possibly cover every detail of the topics in question, so to must software programs need to pick and choose what to document. Indeed, there is no piece of statistical software for which every calculation is described in sufficient detail for it to be replicated.

The book in question is available at most academic libraries, as are the articles that we cite in our manual. The computations are also described in other textbooks like Hamilton. Descriptions of the computation may also be found on the internet via search.

That said, we are always happy to take suggestions on places where the documentation can be improved. Clearly you feel that the VAR chapter needs some work and we will make a note of this for future releases.

nadja123
Posts: 72
Joined: Thu Aug 06, 2009 10:43 am

Re: analytic error bounds for IRFs

Postby nadja123 » Wed Aug 26, 2009 2:39 pm

You know, QMS Glenn, I am not lazy person, I am really trying to learn fast etc, I'm really concerned about details (you may see my latest post under Estimation). I've got a prof who prefers to and is good in computing things and prefers less and is much less good in explaining stuff :cry:
I need to apply stuff completely new to me immediately on a 'research' level :cry: , so sorry for a bit nervous reply :wink:


Return to “General Information and Tips and Tricks”

Who is online

Users browsing this forum: No registered users and 15 guests