Impulse Response standard error bands

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

mofar
Posts: 7
Joined: Wed Aug 18, 2010 12:21 am

Impulse Response standard error bands

Postby mofar » Thu Aug 19, 2010 11:10 am

Hi all:

my question is regarding error bands in impulse responses in EViews. As you know, EViews shows 2 S.D. error band (+,-, 2SE). I want to know if there is any way in EViews to have 1 S.D. error band for impulse responses.

many thanks in advance

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 12462
Joined: Tue Sep 16, 2008 5:38 pm

Re: Impulse Response standard error bands

Postby EViews Gareth » Thu Aug 19, 2010 11:21 am

If you calculate impulse responses from the command line (look up the Impulse View in the VAR section of the Object reference), you can save the standard errors in a matrix (using, say, the smat option), and then you can draw your own graph with as many standard errors as you like.
Follow us on Twitter @IHSEViews

mofar
Posts: 7
Joined: Wed Aug 18, 2010 12:21 am

Re: Impulse Response standard error bands

Postby mofar » Thu Aug 19, 2010 11:34 pm

many thanks for your tips.
Attached is my workfile including 3 annual timse series: oilp (oil prices), gdppc (gdp per capita) and cpi (consumer price index) from 1959-2007. I have followed your recommendation to calculate the response of gdppc and cpi to a one sd shock in oilp:
var var1.ls 1 2 oilp cpi gdppc
var1.impulse(5, m, se=mc, rep=10000, smat=myownimpulse) gdppc @oilp @oilp cpi gdppc

lag numer: 2., IRF for 5 years. Now, I reach to smat option which you said will help to calculate 1 SD error band instead of 2 SE.
in workfile you see the smat output in "myownimpulses". From this stage, I am not sure what should I do to calculate 1 SE band for response of gdppc, cpi to a shock in oilp. Can you please advise me from this stage?
thanks in advance,
Attachments
test_oilp.WF1
impulse response
(13.76 KiB) Downloaded 879 times

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 12462
Joined: Tue Sep 16, 2008 5:38 pm

Re: Impulse Response standard error bands

Postby EViews Gareth » Fri Aug 20, 2010 8:02 am

The vector myownimpulse_se contains the standard errors. Thus a 1SE band is calculated as myownimpulse+1*myownimpulse_se and myownimpulse-1*myownimpulse_se
Follow us on Twitter @IHSEViews


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 52 guests