Dynare forums

Hello everybody,I'm a beginner to use dynare and make dsge model.

I've just made a model like follows:

var zco zy obsy c zc lamd zl l w co2 q zi rk nr p m zm r i ld kd yd pcer ad lc kc ac k yc y pai psd psc g M;

varexo ey ec el ei eco em ead eac eM eg er;

parameters rouy nrss rss rkss ycss ydss wss lamdss kss ldss iss yss css gss pss psdss pscss sigc sigl sigco delta sigm roui rouc roul rouco roum a b cc ebip beta lamdf lamdk lamdl gama rouad rouac rouM roug phim phiy;

model ;
zc-sigc*c=lamd;
zl-sigl*l=lamd+w;
q=-zi;
zco-sigco*co2=lamd+pcer;
q=pai(+1)-(nr/(1+nrss))+(((1-delta)*q(+1)+rkss*rk(+1))/(1-delta+rkss));
zm-sigm*m=lamd+nr/(1+nrss);
zi=roui*zi(-1)+ei;
zc=rouc*zc(-1)+ec;
zm=roum*zm(-1)+em;
zl=roul*zl(-1)+el;
zco=rouco*zco(-1)+eco;

rk+kd(-1)=a*rk+b*w+(1-a-b)*pcer-ad+yd;
w+ld=a*r+b*w+(1-a-b)*pcer-ad+yd;
pcer+co2=a*r+b*w+(1-a-b)*pcer-ad+yd;
rk+kc(-1)=cc*rk+(1-cc)*w-ac+yc;
w+lc=cc*r+(1-cc)*w-ac+yc;
yc=ac+cc*lc+(1-cc)*kc(-1);
yd=ad+a*ld+b*kd(-1)+(1-a-b)*co2;
yss*y=(gama^(1/lamdf))*(ydss^((lamdf-1)/lamdf))*yd+((1-gama)^(1/lamdf))*(ydss^((lamdf-1)/lamdf))*yc;
ad=rouad*ad(-1)+ead;
ac=rouac*ac(-1)+eac;
psd=(1-ebip*beta)*(a*rk+b*w+(1-a-b)*pcer-ad)+ebip*beta*((pss^(1-lamdf))*(pai(+1)+p)+(1-gama)*(pscss^(1-lamdf))*(yc(+1)+yd(+1))/lamdf)/(gama*(psdss^(1-lamdf))+(1-gama)*(pscss^(1-lamdf)))+pai;
psc=(1-ebip*beta)*(cc*rk+(1-cc)*w-ac)+ebip*beta*((pss^(1-lamdf))*(pai(+1)+p)+(1-gama)*(pscss^(1-lamdf))*(yc(+1)+yd(+1))/lamdf)/(gama*(psdss^(1-lamdf))+(1-gama)*(pscss^(1-lamdf)))-(1/lamdf)*(yc(+1)+yd(+1))+pai;
(pss^(1-lamdf))*p=gama*(psdss^(1-lamdf))*psd+(1-gama)*(pscss^(1-lamdf))*psc;

M=rouM*M(-1)+eM;
g=roug*g(-1)+eg;
m=M-p;
k=lamdk*kd+(1-lamdk)*kc;
l=lamdl*ld+(1-lamdl)*lc;
r(+1)=phim*r+(1-phim)*phiy*y(+1)+er;
y=iss*i/yss+css*c/yss+gss*g/yss;
r=nr-pai(+1);
rk=r+delta;
k=(1-delta)*k(-1)+delta*(i+zi);
obsy=y*zy;
zy=rouy*zy(-1)+ey;
end;

But the number of the endogenous variables is more than equations.Which equation should I delete? I think every equation is necessary.What should I do?

Thank you for your reading and answering!

Best regards,

Wu

From what you are saying, you are actually missing an equation that determines the last variable!

jpfeifer wrote:From what you are saying, you are actually missing an equation that determines the last variable!

I'm sorry for my poor English,and which variable am I missing? Thank you.

You will never get an answer here. You are the model builder and the only one knowing how you derived the equations you entered into Dynare. This is not a Dynare problem. If you cannot tell from your model which equation is too much, nobody else can.

jpfeifer wrote:You will never get an answer here. You are the model builder and the only one knowing how you derived the equations you entered into Dynare. This is not a Dynare problem. If you cannot tell from your model which equation is too much, nobody else can.

Thanks very much!! I have solved the problem. But here comes another one.

Error using print_info (line 52)
One of the eigenvalues is close to 0/0 (the absolute value of numerator and denominator is smaller than 1e-6)

Error in check (line 76)
print_info(info, options.noprint);

Error in test2 (line 397)
oo_.dr.eigval = check(M_,options_,oo_);

Error in dynare (line 120)
evalin('base',fname) ;

This problem is made by the parameters or something else?

Use model_diagnostics(M_,options_,oo_) to see that you added a redundant equation and that you did not solve the issue.

jpfeifer wrote:Use model_diagnostics(M_,options_,oo_) to see that you added a redundant equation and that you did not solve the issue.

Amazing!It's useful to me. The model has colinear equations.
Thanks a lot.