Error in computing likelihood for initial parameter values

Estimation ML/Bayesian estimation
1

hello
i have a problem :

please help me

Error in computing likelihood for initial parameter values
??? Error using ==> print_info at 40
Blanchard Kahn conditions are not satisfied: no stable equilibrium

Error in ==> initial_estimation_checks at 69
print_info(info, DynareOptions.noprint)

Error in ==> dynare_estimation_1 at 147
oo_ =
initial_estimation_checks(objective_function,xparam1,dataset_,M_,estim_params_,options_,bayestopt_,oo_);

Error in ==> dynare_estimation at 70
dynare_estimation_1(var_list,dname);

Error in ==> daliri at 228
dynare_estimation(var_list_);

Error in ==> dynare at 120
evalin(‘base’,fname) ;

2

Hi, Before actually launching the estimation of the mode of the posterior distribution, Dynare checks that there is no obvious problem with the initial condition that will be used by the optimizer (used to maximize the log of the posterior kernel). By default, this initial condition (or guess) is the prior mean. In your case the prior mean does not satisfy the Blachard and Kahn condition (you have more unstable eigenvalues than forward looking variables). It is possible to change the default initial condition, using the estimated_params_init block (see the reference manual here). An alternative is to change prior itself, the error may reflect that your prior is problematic.

Best,
Stéphane.

1 Like
3

This can happen even when values for prior mean work when used to simulate the model?

NAME, INITVAL, LB, UB, PRIOR_SHAPE, PRIOR_mean, PRIOR_std_dev. What are the explicit roles of INITVAL and Prior_mean?

4

In those cases, you usually have a unit and are missing diffuse_filter as an option to estimation.

1 Like