I’ve been working all day on this RBC model (attached) and although I’ve tried all sorts of tricks mentioned in this forum I haven’t been able to figure out what’s the deal with the BK indeterminacy issue message here.
I get this message: MJDGGES returns the following error code: 10.
If you skip the check command prior to initval, the error message is: Blanchard Kahn conditions are not satisfied: indeterminacy
I’ve checked the model with the original paper plenty of times and it should work.
Please help. Otherwise I’ll have to switch to Uhlig’s toolkit…
Your problem is the stochastic discount factor. Its timing should always be m instead of m(+1). Also never put the check; command before the initval; and steady command as you are trying to test the BK-conditions at a point where all variables are 0. Only after the steady command do you compute them at the deterministic steady state.
Hello jpfeifer.
The timing on the SDF did the trick. Many Thanks.
However, this brings up the question. If the SDF is by definition a forward looking variable (i.e. t+1) why is it that we need to adjust the timing to contemporaneous for it to work on Dynare?
Please comment and thanks again!
Carlos
Hi,
that is only partly correct. The SDF contains forward looking terms. However, m itself is just a definition of a new name. You want to define something called m. But when you define m(+1), you do not define m, but the conditional expectation of m at time t, i.e. E_t(m(+1)).