occasionally binding constraint binds in steady state

Steady state
1

Dear dynare forum，
i have this equaion x =min( (qstar_s + Z(+1)) * (1 - kh) / d,1), while i want to use perfect foresight simulation. Chatgpt tell me that i should not use max or min function in perfect foresight simulation.
Then i try lmmcp and OCCBIN , like
[mcp=‘x<1’]
x=qstar_s+Z(+1)*(1-kh)/d;
But the problem is that (qstar_s + Z_ss) * (1 - kh_ss) / d_ss>1 in steady state, and thus x=1. In steady state RHS>1, LHS=1 , and ‘The steadystate file did not compute the steady state’.

How can i avoid this problem? i think that all binding occbin constraints in steady state have this kind of problem. Thank you.

2

Have a look at

1 Like
3

What i learn from this mod is that i should not use steady_state_model block. Instead, the combination of initval and steady(solve_algo=10) is the right choice .
Do i get it right?
Thank you a lot, professor Pfeifer!

1 Like
4

Yes, exactly.