I have a fairly big model three-country model, appended. There are two ways to calculate the steady state. One version is usable for symmetric parameterization only, the other version works for asymmetric parameterization, too. The versions can be switched in the file basic_2_steadystate.m by uncommenting and commenting out a few lines as indicated in the file. The model is currently symmetrically parameterized and both versions yield exactly the same steady state, as they should. When I comment out everything in the mod file after “steady”, the programs run fine in either version of the steady state file. But when I uncomment “check” (or uncomment “stoch_simul”) Dynare crashes while using the first version of the steady-state computation. It does not crash for the second version.
Since “steady” is carried out correctly and the results are the same for both versions, I do not understand why “check” or “stoch_simul” has trouble with one, but not the other result.
I upload the files here, but I can upload only 5. Starting values for fsolve are still missing. I upload them in a separate posting. Thanks for looking at the issue.
symm_ss_set.m (10.5 KB)
symm_ss.m (732 Bytes)
basic_2_steadystate.m (842 Bytes)
basic_2_ss.m (17.1 KB)
basic_2.mod (20 KB)