Page 1 of 1
Replication of NK_baseline.mod for Bayesian estimation
Posted:
Fri Jan 30, 2015 2:32 pmby Petryk
Dear Dynare users,
Does anybody have a working version of NK_baseline.mod for Bayesian estimation? I was trying to replicate it using simulated series but there were some issues with parameter 'alppha' in the posterior mode check. I did not observe these issues when I use simulated series in other examples.
Thank you!
Re: Replication of NK_baseline.mod for Bayesian estimation
Posted:
Sat Jan 31, 2015 10:30 pmby costa
Some time ago I wrote this code. Hope it helps.
Re: Replication of NK_baseline.mod for Bayesian estimation
Posted:
Sun Feb 01, 2015 7:57 amby jpfeifer
@Petryk: Could you please post or send the problematic files so I can take a look at them.
Re: Replication of NK_baseline.mod for Bayesian estimation
Posted:
Wed Feb 25, 2015 2:37 pmby Petryk
Dear Dynare experts,
Attached is a working version of NK_baseline.mod modified to replicate the estimation part in Fernández-Villaverde (2010).
While the code runs it shows strange mode checks for ‘alppha’ and some other parameters.
I suspect the problem in the definitions of shocks on line 201-205 but not sure how to fix it.
Also, the definition of shocks in focs has been changed in the original NK_baseline.mod from Fernández-Villaverde and Rubio-Ramírez (2006). To give one example, consider equation (4) in NK_baseline with
‘mu_z(+1)’
compared to ‘z(t)/z(t+1)’ in Fernández-Villaverde and Rubio-Ramírez (2006).
Thank you in advance for help!