Timing of the BGG(2000) model

[firefoxxp]

Dear all, I am confused of the timing of the BGG(2000) model. In the definition of norminal interest rate, BGG(2000) use rn(t+1)=r(t+1)+pi(t+1), and this form works in dynare. I wonder why rn(t)=r(t)+pi(t) does't work in dynare, and what's the difference of the two forms.
Many thanks!

[jpfeifer]

Which paper and which equation are you referring to?

[firefoxxp]

I am sorry. The paper is "The Financial Accelerator in a Quantitative Business Cycle Framework.” Bernanke,Gertler and Griliches(1998)
In section 4.2 "the complete log-linearized model", they defined the norminal interest rate rn(t+1)=r(t+1)+pi(t+1)，and this timing works in dynare,I wonder why the timing rn=r+pi doesn't work.
Many thanks!

[jpfeifer]

My reading is that the notation is misleading.The nominal interest rate is the interest rate between t and t+1 and is actually determined at time t. Basically, this is the Fisher Equation. Hence, it should be

Code: Select all

rn = r + pi(+1);

See https://sites.google.com/site/ambropo/dynarecodes.

[firefoxxp]

Thanks a lot. In the dynare code you mentioned above, the euler equation is c = -r + c(+1), but I think the timing should be c= -r(+1)+c(+1), as the timing in the paper BGG(1999). I wonder why the real interest rate is determined last period.
Many thanks!

[jpfeifer]

I am not that familiar with the model. You might want to ask Cesa-Bianchi directly.

[firefoxxp]

Many thanks!