i wrote a dynare mod file using maximum likelihood estimation,however,the parameters estimation result shows that all the estimated value is equall to the corresponding initiative values,could anyone tell me how to deal with this?
bgg2.mod (1.48 KB)
dsgedata.m (6.33 KB)
I get a crash when I run your MOD file, looks like the maximization ends up on a singular point. Similarly, if I set mode_compute=4, then the maximization convergences, but the result seems spurious (non definite-positive hessian at the maximum).
I would suggest to modify your initial values, so that you have a better chance of finding a non-spurious result. If that does not help, I guess the only choice is to move to Bayesian estimation, and assign priors…