Question regarding stack_solve_algo

Dear all,

I have a question regarding the different solution algorithms used for perfect foresight simulations. I have a model, which I have written with an exogenous steady state file. The model’s perfect foresight solution can be found using the default algorithm as well as the option ‘stack_solve_algo=7’.
steady state (4.9 KB)

However, once I solve an equivalent model without the steady state file, but with an initval_file, I find that the default algorithm is no longer able to find the solution. The option ‘stack_solve_algo=7’ does find the solution that is the same with the one solved with the external steady state file. I’m quite baffled by this and would like to ask what went wrong with my initval_file, as I plan to extend on the current model and it seems weird that it’s not compatible with the default algorithm. My intuition is that the init_val file does not solve the steady state, which model_diagnostics report to be the case. Would that be the reason why the default algorithm is not working here?
initval (5.4 KB)

Best regard,

From what I can see, the two experiments are not the same. Inspecting oo_.exo_simul after perfect_foresight_setup shows for the init_val_file-version
while for the other one
Thus, the shock sequence is different.

Thanks jpfeifer! I did not recognize the difference before. I’ll double check on my code.