Errors in computing likelihood and Blanchard-Kahn condition

Estimation
1

Dear community,

I hope this message finds you well. I am a university student from Korea, currently studying macroeconomics. Since being introduced to DSGE models in the last semester, I have been eager to apply these models in practice.

I have constructed a model and attempted to estimate the parameters using a dataset I compiled. However, I have encountered several errors, including “Error in computing likelihood for initial parameter values” and “Blanchard & Kahn conditions are not satisfied: indeterminacy,” which have hindered my progress.

I would be extremely grateful if you could kindly review my model and guide me on what might be causing these issues.

Additionally, I apologize for posting my question in the “Dynare Help” category without fully understanding the community guidelines.

Thank you very much in advance for your assistance.

Very Respectfully,

Mac Park

wow.mod (2.6 KB)
datafile.xlsx (15.5 KB)

P.S. The values in the data file represent growth rates obtained through log differencing.

2

From a quick look:

  1. Before attempting estimation, it is always best to run stoch_simul to ensure that the model will actually solve. Your model doesn’t and you have many undefined parameters (so they will be automatically set to zero, which may be the source of the error).
  2. Your data is not mean-zero but I think that your model observables are, producing an inconsistency.
    Once your model solves under stoch_simul you can then move on to estimation.
3

Thank you very much for your valuable response. I will take your advice into consideration and give it another try. Once again, thank you for your help.

However, I have an additional question. Is it possible to estimate the remaining parameters after setting the initial values for some parameters? I am still a bit confused about the estimation method, so I apologize for asking again.

4

Yes, that is fine. You do not need to estimate every single parameter in your model.

5

Thank you for your great advices!

Have a nice day :slight_smile: