Svar_identification block of the MS-SBVAR model

BVAR
1

Hello everyone,

I’m really struggling with the svar_identification block of the MS-SBVAR model. The Dynare manual states that the code block will be documentated, however, the wiki to which I’m directed to is no longer available which is really frustrating. I would really appreciate any kind of help with regard to the explanation of this code block. Thanks in advance.

Block:svar_identification ;

This block is terminated by end; and contains lines of the form:

UPPER_CHOLESKY;
LOWER_CHOLESKY;
EXCLUSION CONSTANTS;
EXCLUSION LAG INTEGER; EQUATION INTEGER, VARIABLE_NAME [,VARIABLE_NAME...];
RESTRICTION EQUATION INTEGER, EXPRESSION = EXPRESSION;
To be documented. For now, see the wiki: https://archives.dynare.org/DynareWiki/MarkovSwitchingInterface
2

The wiki problem seems to be a server issue. We are working on it. For now, have a look at the example files at tests/ms-sbvar · master · Dynare / dynare · GitLab

1 Like
3

Dear Professor,

Thank you so much for your help. I have spent time going through the example files and the wiki issue have been solved which really helped in clarifying a lot of the confusions I had. If you dont mind me asking. How do I get the confidence interval around the ms_irf() function? Thanks in advance.

4

Which type of uncertainty are you looking for? There is the parameter_uncertainty option. See tests/ms-sbvar/test_ms_variances.mod · master · Dynare / dynare · GitLab

1 Like
5

I’m trying to get similar results (Figure 5) obtained by Stéphane Lhuissier in his paper “Financial Intermediaries’ Instability and Euro Area Macroeconomic Dynamics.” (link: http://stephanelhuissier.eu/assets/ea-intermediaries-v2.0.pdf). The author obtained the 68% error bands of the impulse response functions. The ms_irf function has the option error_band_percentiles, when setting the latter equal to [0.16 0.50 0.84] I don’t get the error bands as in the paper. The parameter_uncertainty gives indeed a confidence interval, I just don’t know at what risk level (is it 1%, 5% or 10%). I would really be forever grateful for your kind help Professor Johannes.

6

What do you mean with

? Are no bands at all shown? My guess is that parameter_uncertainty uses the parameter_uncertainty you specify.