Problem with estimating on empirical data

Estimation ML/Bayesian estimation
21

  1. It was a tipping error as I have tried to include/exclude the equation and then missed a log - that does dot change the error message unfortunately.

  2. I follow your point, but i think it is because i’m not quiet sure how to include the extra shock to c_obs relative to the other model equations so it is observed by other parts of the model?

Thanks in advance .

22

The error message comes from point 2. If a parameter is not identified, the likelihood will be flat in this direction and the Hessian not positive definite. You can run identification to see this.