Bayesian VAR

[EViews Gareth]

The BVAR add-in doesn't support non-simple (i.e. 1-7) lag structures.

[mjfl]

The BVAR add-in doesn't support non-simple (i.e. 1-7) lag structures.

hi Gareth,
thks for the reply.

but can anyone teach me where should i change or tinker from becasue i cant seem to get the lag lag correct in program code and sub routines. can i have a look of how VAR source code looks like to change the BVAR one?

tyrying to make it more operational. Im gg from Matlab to Eviews with all these. cheers to developers.

[EViews Gareth]

VARs are not add-ins like BVARs - they are not written in an EViews program.

[mjfl]

VARs are not add-ins like BVARs - they are not written in an EViews program.

hi Gareth,

thank you once again. that i know. just wondering is there a way i can get hold of VAR source code? c-lang?
or how can i finetune BVAR program code?
much much appreciated.

[EViews Gareth]

The VAR code would not be useful to you.

[cobleton]

Dear Donihue,

The IRF graphs display is too small to read text characters when several variables are used. Would it be possible to have an option to show the IRF graphs by equation?

The smart EViews gives you an option, "Extract selected graph", by right-clicking on the IRF graphs.

And whilst still on this subject (sorry!): is there any chance of having the IRF graphs display a +/- 1 (and/or 2) std. error band?

I will put it on my to-do list.

Best,
Esther

Dear Esther,

Just a question about IRF confidence intervals. Is it straightforward to build them? Is it possible to have an update of the Add-in which also includes the bands? This would be extremely useful (as already the BVAR add-in is) for any central banker/ graduate student.

Thanks

[Keneni09]

Dear all,

I am so greatful if anyone of you give me a hint on the following issues with respect to using the add-in for BVAR
1. The Sims-Zha Normal-Wishart allows for the exogenous variables. The impulse responses are only for the endogenous variables.
Is there a way that I get the responses of the endogenous variables as a result of the changes in the exogenous variables?
2. Is it possible to recover the variance decomposition after the estimation?

Thank you so much for your help in advance.

Keneni

[EViews Esther]

None of them are available.

[JakeS]

I have not been able to get the bvar add in to work at all. I continually get the error '_SH01' is not defined. I have tried changing the estimation period, the variables I used, normalizing the data, different priors but nothing seems to work.

[EViews Gareth]

Might be useful to provide your workfile.

[JakeS]

Here is what I am working with now.

[EViews Gareth]

And details of what you've done...

[JakeS]

And details of what you've done...

Installed the latest version of the BVAR add-in, from the link on the first page of this thread (Eviews v7.2)
Imported some data from excel (real gdp , cpi, civilian employment, industrial production, etc) 1980-2012q1

clicked 'add-ins' -> Bayesian Var

Entered the variable names in the endogenous section plus the following:
estimation period=1980q4 2011q4, lags = 4, calculate irfs, Sims-Zha Normal-Wishart
defaults on everything else.
Clicked 'ok'
get the screen with the sim-zha hyperparameters, just chose defaults.
Clicked 'ok'

I then just get an error described above

[EViews Esther]

What are your endogenous variables? The error message indicates that there exists a near singularity problem.

[Stefania A.]

I think I’m having a similar problem.
I attached my workfile (work1 is after I did all the steps below, work is before).
I want to estimate a BVAR with the variables m, n, f, k, l and i with one lag, using the Minnesota/Litterman prior (and calculate impulse responses for 10-20 periods).

So, after installing the BVAR add in and estimating a regular VAR, I opened the BVAR and wrote:
Endogenous variables: m n f k l i
(just like this, no commas)
Exogenous variables: I wrote nothing
Estimation sample 1990 2010
(it appeared as default because of the workfile specification, I suppose)
1 lag
Calculate impulse responses, 10 periods
Koko Minnesota/Litterman prior
Name: stef
The next dialog box appeared, I left everything as default and after clicking OK, it said “Syntax error” and a lot of new things (series, matrixes, a group) appeared in my workfile…

Could you please help me…
Thank you so much in advance!
(Needless to say this is part of project that I have to submit this Saturday…)