I am trying to simulate the extended model of Iacoviello (2014) - for his paper titled Financial Business Cycles.
I have created a bare version of the model, which should check out, but after several attempts, I have encountered an error message which reads: “An infinite element was encountered when trying to solve equation(s) 27 with respect to the variable(s): epss.”
How do I correct this and get my steady state results? Could it be that there is a problem with my initial values?
Please help! Attached is my “.mod” file and the paper.
I would say your shock equations are wrong. Look at e.g.
exp(az)=rhoaz*exp(az(-1))+vz;
For the rhoaz not equal to 0 the only steady state would be exp(az)=0, but this is impossible as it would imply az=-Inf. Either drop the exp() in these equations or rewrite them in the correct way as e.g.
exp(az)=exp(az(-1))^rhoaz*vz;
Note that this inconsistent log substitution you performed also affects the other equations of the model were the shocks enter. You need to define them consistently. This is the reason the error message appears.
I made the changes to all 4 shock equations as you recommended. The model still won’t compute the steady states. It is still showing the same error message:
“An infinite element was encountered when trying to solve equation(s) 27 with respect to the variable(s): epss.”
Shock equations were changed as recommended, to:
%Shock Equations (30-33):
exp(epse)=(exp(epse(-1)))^rhobeve; //% Shock to repayment of subprime guy
exp(epss)=(exp(epss(-1)))^rhobhvs; //% Shock to repayment of subprime guy
exp(ak)=exp(ak(-1))^rhoakvk;
exp(az)=exp(az(-1))^rhoazvz;