Hello Guys, I am trying to replicate a paper. Is a bayesian estimation of a linearized DSGE model. I receive the following error message. “Error in computing likelihood for initial parameter values.” I am not sure why is giving this error, I have the same model calibrated with the priors that I am using for the Bayesian estimation and it works. Is there anything wrong on trying to make a bayesian model that is already linearized? PS: sorry I don’t know how to upload my data file. I uploaded a xls file
Error in computing likelihood for initial parameter values
Error using print_info (line 40)
Blanchard Kahn conditions are not satisfied: no stable equilibrium
Error in initial_estimation_checks (line 69)
Error in dynare_estimation_1 (line 169)
Error in dynare_estimation (line 70)
Error in dsge_bubbles_b (line 373)
Error in dynare (line 120)
data.xls (56.5 KB)
dsge_bubbles.mod (5.51 KB)
dsge_bubbles_b.mod (6.81 KB)
You must provide the caibrated values as explicit starting values. Otherwise, Dynare uses the prior mean.
Thanks Jpfeifer, from the beginning I provided the calibrated values as explicit starting values, but it does not work.
No you don’t. You initialize the parameters, but you don’t provide those initial values as starting values in either the estimated_params-block nor do you provide an estimated_params_init-block after the estimated_params-block.
Ok I see, I included this model block after the estimated_params block, but still is not working. Also is my understanding that if do not include these block or initial values at the estimated_params block it will use the prior means as initial values, please let me know if I am wrong. I am new in dynare
stderr e_a, 0.01;
stderr e_am, 0.01;
stderr e_z, 0.01;
stderr e_zm, 0.01;
stderr e_psi, 0.01;
stderr e_xi, 0.01;
stderr e_theta, 0.01;
The data you posted is identical for all variables.