Dynare forums

Hi there,
I just downloaded Dynare 4.2.0 and I am experiencing some errors with a very simple model that used to work smoothly in Dynare 4.1.3

The model is a sinmple version of Clarida, Gali, Gertler (1999) augmented with am exchange rate equation. If the exchante rate is defined as a first difference (so that it is stationary) no problem arises. But if I define the exchange rate in levels (and therefore use the unit_root_vars command) it crashes in Dynare 4.2.0 (but not in Dynare 4.1.3)

The error i get in version 4.2.0 is as follows:

Error in computing likelihood for initial parameter values
??? Error using ==> print_info at 39
Blanchard Kahn conditions are not satisfied: no stable equilibrium

Error in ==> initial_estimation_checks at 101
print_info(info, options_.noprint)

Error in ==> dynare_estimation_1 at 367
initial_estimation_checks(xparam1,gend,data,data_index,number_of_observations,no_more_missing_observations);

Error in ==> dynare_estimation at 62
dynare_estimation_1(var_list,varargin{:});

Error in ==> modelo_css at 196
dynare_estimation(var_list_);

Error in ==> dynare at 132
evalin('base',fname) ;

Error in ==> dynare_run at 6
dynare modelo_css.mod

Any ideas?
Thanks in advance,
Ezequiel

Hi,

Your problem comes from the fact that your model has a unit root, and that you are using the "normal" Kalman filter.

Previously, Dynare was accepting this, even though it is not correct. Now, Dynare refuses to do so, even though I admit that the error message is not very explicit.

The solution is to use the diffuse Kalman filter, by adding the "diffuse_filter" option to estimation command.

Best,

Thanks Sebastien,

It now works. Thanks for the promt reply.
Cheers,
Ezequiel

ezequielac wrote:Thanks Sebastien,

It now works. Thanks for the promt reply.
Cheers,
Ezequiel

could you share your final version?