steady state command

[jumo]

Hello!

Can someone please tell me the difference between these two steady state commands:

steady;
steady(solve_algo=0);

For example when estimating the Gerali et al (2010) model I get different steady state values for inflation. Then when I want to estimate the so called "FF" version the first command does not find the steady state anymore but the second one does...is one command in general more precise or preferable?

Thank you!

[jpfeifer]

Code: Select all

steady;

uses a different algorithm than

Code: Select all

steady(solve_algo=0));

For unique steady states they should find the same result (apart from small numerical differences in the order of 1E-6). If (some elements of) the steady state are not uniquely determined, you may end up with different values. Inflation and the nominal interest rate in New Keynesian models are typically not endogenously determined and you need to fix one of the two to get the other one. In this case, different mode-finders may end up with different values.

And no, you cannot say which one performs better.

Lastly, try running model_diagnostics on your model.