Dynare forums

Dear Everyone:

We have two country models, could we have a model with two firms with different producion function?
I tried, but there are problems of "empty state space model":


var y_h y_f c_h c_f k_h k_f i_h i_f l_h l_f w r z_h z_f;
varexo e_h e_f;
parameters beta psi delta alpha_h alpha_f rho_h rho_f;

alpha_h = 0.33;
alpha_f = 0.53;
beta = 0.99;
delta = 0.023;
psi = 1.75;
rho_h = 0.95;
rho_f = 0.98;

model;
1/c_h = beta*(1/c_h(+1))*(1+r(+1)-delta);
1/c_f = beta*(1/c_h(+1))*(1+r(+1)-delta);
psi*c_h/(1-l_h) = w;
psi*c_f/(1-l_f) = w;
c_h+ k_h-(1-delta)*k_h(-1)= y_h;
c_f+ k_f-(1-delta)*k_f(-1) = y_f;
y_h = (k_h(-1)^alpha_h)*(exp(z_h)*l_h)^(1-alpha_h); // h: production
y_f = (k_f(-1)^alpha_f)*(exp(z_f)*l_f)^(1-alpha_f); // f: production
w=y_h*(1-alpha_h)/l_h; // h. wage, perfect competition
r=y_h*alpha_h/(k_h(-1));
w=y_f*(1-alpha_f)/l_f; // f.wage, perfect competition
r=y_f*alpha_f/(k_f(-1));
//i_h = k_h-(1-delta)*k_h(-1);
//i_f = k_f-(1-delta)*k_f(-1);
//y=y_h+y_f;
//l=l_h+l_f;
//k=k_h+k_f;
//i =i_h+i_f;
z_h = rho_h*z_h(-1)+e_h;
z_f = rho_f*z_f(-1)+e_f;

end;

initval;
k_h = 8;
k_f = 8;
//k = 16;
c_h = 0.76;
c_f = 0.76;
l_h = 0.3;
l_f = 0.3;
//l = 0.6;
w = 2.07;
r = 0.03;
z_h = 0;
z_f = 0;
e_h = 0;
e_f = 0;
end;

steady;
check;
shocks;
var e_h = 0.1;
end;

stoch_simul(irf=40,periods=2100);

Your help is highly appreciated!

Best,

Yi

My Dynare version warns that it cannot find the steady state. The reason is that your starting values are really poor. Put

Code: Select all

resid(1);

before steady to see this.

Dear Jpfeifer,

Thank you very much for the help!

I guess the two wage and rent collide with each other, so"Empty state-space model" appears.

Is there anyway to build a heterogeneous firm model in Daynar using this simple way, not with the distribution of firms?

Thanks again!

Best,

Yi