Let’s say that I am going to estimate my model and my steady state values are calculated like this
model;
pk_ss=gamma/beta+delta-1;
x_ss=pie;
p_flex=1;
inflation_ss=pie;
mc_ss=(eta_p-1)/eta_p;
Mw_ss=eta_w/(eta_w-1);
R_ss=gamma*pie/beta;
h=1;
mu=0;
PSI_ss=((1-alpha)/((pk_ss/alpha*gamma^(1/omega)*mc_ss)^(omega-1)-alpha/(gamma^((omega-1)/omega))))^(omega/(omega-1));
TETHA_ss=1/(gamma^((1-xi)/xi));
PHI_ss=(alpha*(PSI_ss/gamma)^((omega-1)/omega)+(1-alpha))^(1/(omega-1));
w_ss=(1-alpha)mc_ssPHI_ss;
OMEGA_ss=(TETHA_ssMw_ss(1-theta)^sigma)((PHI_ss^omega)(1-sg)-(gamma-1+delta)*PSI_ss)^sigma;
n_ss=(w_ss/OMEGA_ss)^(1/(sigma+phi));
k_ss=n_ss*PSI_ss;
investment_ss=k_ss*((gamma+delta-1)/gamma);
y_ss=(alpha*(k_ss/gamma)^((omega-1)/omega)+(1-alpha)*n_ss^((omega-1)/omega))^(omega/(omega-1));
further assume that I have a Taylor rule in my model block that reacts on output deviations from its steady state according to
R=R_ss*(y/y_ss))^rho;
This means that I want to include my steady state value for y in my model block and I also want it to be updated as my parameters change during the estimation and as was evident from the steady state ekvations above my steady state value for y_ss depends on a lot of expression equations.
Then the problem is. What do I write under the “steady_state_model;” block? Because I can’t use the same expression names like THETA_ss and it seems a bit messy to have to define new expression variables just for the steady state block such as “THETA_ss2”. So does anyone have a clever way how to include all of this such that all steady state equations are updated during the estimation in a way that I can include “y_ss” directly in the model?