Dear all, I have a DSGE model in which the parameters are all calibrated. The model ends by stochastic simulation instead of estimation. This model is able to run successfully. However, when I try to modify the model in to estimate some parameters instead of calibrating them, the modified model can’t work. The error messages are as follows.

Error using initial_estimation_checks (line 153)

initial_estimation_checks:: The forecast error variance in the multivariate Kalman filter became singular.

Error in initial_estimation_checks (line 153)

error(‘initial_estimation_checks:: The forecast error variance in the multivariate Kalman filter became singular.’)

Error in dynare_estimation_1 (line 164)

oo_ = initial_estimation_checks(objective_function,xparam1,dataset_,dataset_info,M_,estim_params_,options_,bayestopt_,bounds,oo_);

Error in dynare_estimation (line 105)

dynare_estimation_1(var_list,dname);

Error in estimate_phi.driver (line 395)

oo_recursive_=dynare_estimation(var_list_);

Error in dynare (line 293)

evalin(‘base’,[fname ‘.driver’]) ;

Could you help me solve the problem? Thanks!

P.S. “stochsimul.mod” is the original model. “stimate_phi.mod” is the modified model. “inochini.m” is the datafile.

estimate_phi.mod (2.2 KB)

stochsimul.mod (1.9 KB)

inochini.mat (2.6 KB)