I am trying to compute the steady state for my model, but I encounter a small residual in one of the equations (around the sixth decimal place). I have tried different solving algorithms (except for the homotopy method with varphi), but I still cannot eliminate the residual completely.
If anyone believes that using homotopy (or any other approach) might help, I would really appreciate it if you could enlighten me or guide me on how to proceed.
Alternatively, if solving for the exact steady state is too difficult, I would like to ask:
Is there a recommended way to compute the IRFs even if one equation still has a small nonzero residual?
I am not sure I understand the problem. There is no error message related to the steady state. The small residual is within the tolerance of numerical solver used. Thus, there is no issue. You can simply run stoch_simul.
Dear Professor Pfeifer, thank you for your message. After running the irf.mod file, I have the following error:
Error using print_info (line 33)
The generalized Schur (QZ) decomposition failed. For more information, see the documentation for Lapack function dgges: info=30, n=16. You can also run model_diagnostics to get more
information on what may cause this problem.
Error in stoch_simul (line 119)
print_info(info, options_.noprint, options_);
Error in irf.driver (line 917)
[info, oo_, options_, M_] = stoch_simul(M_, options_, oo_, var_list_);
Error in dynare (line 308)
evalin(‘base’,[fname ‘.driver’]);
I was under the impression that this was due to the residual in the steady state, but perhaps it is a different issue. Do you have any recommendation based on this error?
Thank you in advance.
Indeed, that is not about the steady state, but usually about the timing in the model equations are entering a redundant equation and therefore missing an important one.