Page 1 of 1

montenearsvar.rpf

Posted: Fri Sep 08, 2017 1:15 pm
by TG81
1. I am using the RATS program montenearsvar.rpf to generate impulse responses for my VARX model. I want to generate impulse responses due to 1% shock in each variable. How can I do that within the program?

2. I want to generate the FEVD's for the same model. For that I can use the errors instruction after the cvmodel in the montenearsvar program or I can use @mcfevdtable. However, the answers do not match for the point estimates of mcfevdtable and what we get from the error instruction. What could be the reason? What is the correct way of calculating VDFs in such a model?

Thanks in advance!

Re: montenearsvar.rpf

Posted: Fri Sep 08, 2017 9:58 pm
by TomDoan
TG81 wrote:1. I am using the RATS program montenearsvar.rpf to generate impulse responses for my VARX model. I want to generate impulse responses due to 1% shock in each variable. How can I do that within the program?
If you're doing an SVAR (Structural VAR), 1% shocks to each variable make no sense. MONTESUR is a (much) simpler program for doing Cholesky or other constructed shocks.
TG81 wrote: 2. I want to generate the FEVD's for the same model. For that I can use the errors instruction after the cvmodel in the montenearsvar program or I can use @mcfevdtable. However, the answers do not match for the point estimates of mcfevdtable and what we get from the error instruction. What could be the reason? What is the correct way of calculating VDFs in such a model?
Please read the description of @MCFEVDTABLE.