Dynare forums

Dear Experts,

Attached please find a very simple two-country model, which has a bond as financial intermediary. I use it to test the response of current account when the two country have asymmetric intertemporal elasticity of substitution, when facing symmetric TFP shocks.

we have the following equations:

c1^(-gamma1)=beta*c1(+1)^(-gamma1)*(1+Rb(+1));
c2^(-gamma2)=beta*c2(+1)^(-gamma2)*(1+Rb(+1));

b1+c1+i1= y1+b1(-1)* (1+Rb);
b2+c2+i2= y2+b2(-1)* (1+Rb);

I was wondering if I should write Rb(+1) in equation 12 and Rb in equation 34, or, use Rb and Rb(-1) instead? In fact, if I write Rb(+1) and Rb, the code does not work, and report "Matrix is singular..." .

Could you please help me?

Thanks a lot!

Without knowing your model, this is hard to answer. But considering that your timing uses b1(-1) I guess you did not use the predetermined variables command. I don't know how the interest rate is determined in your model, however usually in those models people say it is contractually agreed one period in advance. Then it should be Rb and Rb(-1).

Professor, my model is very standard and simple two-country model, attached please find the note. I wonder if there is any theoretical problem with writing Rb(+1) in the bond Euler equation and Rb in the budeget constraint? Since for the interest rate for capital we have MU_C(t)=beta*MU_C(t+1) *R(t+1), which is parallel to the bond Euler euqation. But the problem is, when I use Rb(+1), the model can not work and reports singular. I was wondering why?

Thanks a lot!

jpfeifer wrote:Without knowing your model, this is hard to answer. But considering that your timing uses b1(-1) I guess you did not use the predetermined variables command. I don't know how the interest rate is determined in your model, however usually in those models people say it is contractually agreed one period in advance. Then it should be Rb and Rb(-1).