Gali & Monacelli 2005

General DSGE Modeling Replication Attempts
1

Dear Mr. jpfeifer,

I’ve downloaded your code Gali_Monacelli_2005.mod, but I was wondering why there is no Omega in equation (35):
ynat = Gamma*a + alpha*Psi*ystar
and no rho in equation (37):
rnat = -sigma_a*Gamma*(1-rhoa)*a + alpha*sigma_a*(Theta+Psi)*(ystar(+1)-ystar).

I thought it was to simplify the calculation of the steady state of the model , incorporating constant terms in some definitions. So I’ve tried to written these equations using rho and Omega with the initval command.
Dynare couldn’t find the steady state and I suspect that there is some other equation that I have to adjust to do that. But despite I tried to open all the calculations. In particular, I suspect that equation (35) (natural level of output) is not consistent with equations (22) and (27), since the output is not around Omega.
Would you help me, please?

2

Usually, you either need the constants in all equations or none. I decided to go for leaving out the constants as not all required constants were given in the paper.

3

Thanks for your answer, professor.
Just one more question: are these constants relative to approximations in log-linearized equations or maybe is there some other reasons? For example, I can’t see why rho could be excluded from the natural rate of interest since beta<1.

4

These constants are related to the steady state. If you leave them out, you have everything in deviations from steady state.

5

I’ve got it! Thank you so much, professor.