Dear all,

I’m approaching the estimation with dynare, and I’m encountering some difficulties.

I tried to simulate a small model and then to estimate its parameters via maximum likelihood using 3 series obtained by the simulation.

If I slightly move the initial value of the parameters from the “true” one, estimation fails: estimates are far

from the peaks in the mode_check plots and, if I check the residuals, I receive the following message:

Residual for equation number 1 is equal to NaN

Residual for equation number 2 is equal to NaN

Residual for equation number 3 is equal to NaN

Residual for equation number 4 is equal to NaN

Residual for equation number 5 is equal to NaN

Residual for equation number 6 is equal to 0

Residual for equation number 7 is equal to NaN

Residual for equation number 8 is equal to NaN

Residual for equation number 9 is equal to NaN

What should I check?

Many thanks for any suggestion.

Best,

Andrea

UPDATE: I found my mistakes:

- I didn’t declare parameters values before the model, thus I got NaNs;
- I didn’t compute the analytical steady state.