(1) When I use the ramsey_policy command in Dynare, does it compute the optimal policy under commitment or are the policies time consistent or are the policies optimal in a timeless perspective?
(2) In the document "DSGE Models with Dynare++. A Tutorial" by Ondra Kamenk, it is said that "Dynare++ solves the optimal policy problem with timeless perspective". If there are forward looking constraints then there will be some backward looking multipliers. I think Dynare++ doesn't solve the optimal policy in a timeless perspective, since if you run the example in the document in Dynare++ and look at the dump file, you will find some backward looking Lagrange multipliers. As laid out in the document you will have to set the backward looking multipliers to 0 and again simulate the model. See the following code from the document.
- Code: Select all
>> load kp1980_2.mat
>> shocks = zeros(1,100);
>> ystart = dyn_ss;
>> ystart(dyn_i_MULT3) = 0;
>> r=dynare_simul('kp1980_2.mat',shocks,ystart);
So am I right to think that whenever you do an optimal policy in Dynare++, you will have to look at the dump file for any backward looking Lagrange multipliers and set it to 0 ? By setting the backward looking multiplier to 0, can we say that we have optimal policy under commitment in a timeless perspective?
By the way why do u have to set the both the shock and multiplier to 0 ? Is it wrong if we set only the multiplier to 0 with the shocks remaining as it is?
(3) I ran the same example from the document in both Dynare and Dynare++ (without setting the backward looking Lagrange multiplier to 0) and compared the IRFs. The IRFS from Dynare and Dynare++ are same. Does that mean the "ramsey_policy" command doesn't solve the optimal policy under commitment in a timeless perspective? I am confused since Dynare says that it computes ramsey policy under commitment and the IRFs computed with Dynare++( and backward looking multipliers still present) are same as IRFs from Dynare.
(4) Does Dynare or Dynare++ solves Ramsey policy when there are 2 instruments or in other words, 12 equations and 10 variables?
I will be extremely grateful if anyone can explain my doubts.
Thanks!!
Sub_macro