The RATS Software Forum

Dear Tom,

There is another question about DSGE-VAR. To my understanding, DSGE-VAR is that DSGE model has been solved via DSGE instruction (in RATS) to obtain matrix A and F in state-space form. The next step is to use procedure varfromdlm to have the corresponding VAR model. Is it right?

Thanks in advance,

I've seen the phrase used in at least two ways---either as you describe (which generates the VAR coefficients from the covariance generated by a DSGE solved out to state-space form) or as a VAR estimated from the data with a prior derived from a DSGE.

Dear Tom,

Thanks for your reply. Could I have an example code for any DSGE-VAR model (an replication work) that you already have ?

Thanks in advance,

To be honest, I don't have any.