No Posterior after Bayesian estimation of DSGE model

Estimation ML/Bayesian estimation
1

Dear,

I am trying to estimate a DSGE model with unemployment. I use a code inspired from Smets and Wouters (2003)'s code but with unemployment in order to replicate Gali, Smets and Wouters (2012) for the Euro Zone.
The model seems to work.
However, the Bayesian estimation is not computed, I obtain no posterior (or only wrong ones with zero stdeviation).
I tried to with several mode_compute, in particular the mode_compute = 6 and I tried with several initial values and priors.

You will finds attached the data and the mod file.

Do you have an idea of what could be wrong?

Thank you,

GSW.mod (10.2 KB)
base.xls (39.5 KB)

1 Like
2

1.Your data is not read in correctly, because the first column of your Excel files does not have a column header.
2. You are not handling parameter dependence correctly. Some of your parameter are not identified.

    pi_bar is not identified in the model!
    [dJ/d(pi_bar)=0 for all tau elements in the model solution!]
    tau_bar is not identified in the model!
    [dJ/d(tau_bar)=0 for all tau elements in the model solution!]
    beta_const is not identified in the model!
    [dJ/d(beta_const)=0 for all tau elements in the model solution!]