# Impossible to find steady state?

Hi,

I’ve built a NK Model with banks. I’m unable to run the model. The error message I’m getting is:

Impossible to find the steady state (the sum of square residuals of the static equations is
11676028617952258251233716862976.0000). Either the model doesn’t have a steady state, there are an
infinity of steady states, or the guess values are too far from the solution.

I can’t seem to figure out why my sum of resid is so huge?

This is the code:

``````var ir c l y w r iq q d f i m z id o;
varexo eps_e;
parameters ETA ALPHAM SIGMAM ALPHAD SIGMAD ALPHAR SIGMAR THETA BETA GAMMA RHOR RHOI RHOY CHI;
ETA = 2.7;
ALPHAM = 0.175;
SIGMAM = 0.95;
ALPHAR = 0.125;
SIGMAR = 0.95;
THETA = 0.75;
BETA = 0.9823;
GAMMA = 0.18;
RHOR = 0.8;
RHOI = 1.2;
RHOY = 0.5;
CHI = 0.95;
STDERR_AE = 0.0029;

model;
1/exp(c) = BETA*(exp(ir)/(exp(c(+1))*exp(i(+1)))) ;
exp(w)/exp(c) = (exp(l)^(ETA)) ;
1/exp(c) = ALPHAM/(exp(m)^SIGMAM) + BETA*(1/(exp(c(+1))*exp(i(+1)))) ;
1/exp(c) = ALPHAR/(exp(r)^SIGMAR) + BETA*(1/(exp(c(+1))*exp(i(+1)))) ;
exp(iq) = (1/BETA)*(exp(f(+1))/exp(f)) ;
exp(id) = (1/BETA)*(1 - GAMMA)*(exp(f(+1))/exp(f)) ;
exp(w)*exp(l) = exp(z)*exp(y)/(1 + ((exp(iq))*(1 + exp(iq)))) ;
i = BETA*i(+1) + (((1 - THETA)*(1 - (THETA*BETA)))/THETA)*z ;
exp(o) = CHI*exp(o(-1)) + eps_e ;
exp(y) = exp(l) ;
exp(y) = exp(c) ;
(1 + exp(id)) = (exp(q)*exp(w)*exp(l))/exp(d) ;
exp(f(+1)) = (exp(f)*(1 + exp(iq))) + (exp(d)*(exp(iq) - exp(id))) ;
exp(ir) = (exp(ir(+1))^(RHOR))*((exp(i)^(RHOI))*(exp(y)^(RHOY)))^(1 - RHOR) + exp(o);
end;

initval;
ir = 1.02;
c =0.01614;
l = 0.01614;
y = 0.01614;
w = 0.7501;
r = 0.090477;
id = 0.8364;
iq = 1.02;
q = 36.2521;
d = 0.0062787;
f = 0.010013;
i = 1;
m = 0.12497;
z = 1.0404;
o = 0;
end;

shocks;
var eps_e; stderr 100*STDERR_AE;
end;

stoch_simul(dr_algo=0,order=1,irf=20) y i ir q c;
``````

Decided to keep the equations in levels for simplicity.

I’ve kept my equations in levels (except for the NKPC). I’ve simplified the model, gone over it multiple times, ensured my functional forms are correct, parameter values are correct, derived the SS values from a combination of doing it by hand and using fsolve on matlab. I can’t see what’s wrong with my model and what to do about it.

Would appreciate any help, thank you!

Only do the exp substitution once the model works. For example, your Taylor rule has `+ exp(o)`, which will be `+1` in steady state.

Tried that. Still can’t find the steady state. The sum of resids is smaller, but still very large

Then you need to continue searching for mistakes like the one I outlined.