Question about prior and observations

Hi everyone. I try to replicate the Milani (2016) paper( The purpose of this paper is that showing the DSGE model does not match the expectation data by DSGE-VAR approach.

I have two questions.

  1. In this paper, one of the parameter’s prior distribution is set to 100*xi~Gamma(0,15,0.05). How I declare in estimated_params block?

2 This paper uses actual data that output gap, inflation rate, and fed funds rate (x_obs,pi_obs,i_obs)
from FRED and expectation data that output and inflation (exp_x_obs, exp_pi_obs) from SPF. All of the data are seasonally adjusted.
I declare these observations to

x_obs=x ;

Is this correct? Since this is my first time to treat expectation data, I do not know how to declare expectations.

Best regards.

Hi, For the parameter you can either change the prior mean and prior variance (I don’t know what are the parameters in your Gamma function) or change the way you input the parameter in your model (replacing \xi by \xi/100 in the equations). Regarding the second question, I would do it as you said but I didn’t read the paper.


Thanks to quick reply!
Gamma(0,15,0.05) is a just typo of Gamma(0.15,0.05). Sorry.

I try to estimate right now.
Sincerely yours.