Hello everybody!
I would like to compute the Baxter and King filter of output in the model section treating it as an endogenous variable, say Y_BK. Is it possible to do something like this in dynare?
From what I can see, the Baxter/King filter has a state space representation that you could use to manually define the filtered output in your model.
Where can I find the formula for its state-space representation?
Their paper shows how to compute the moving average weights.
So, I computed them and them used stoch_simul; but the IRFs I get look strange:
Attached you can find my .mod file.
NK_Model.mod (6.9 KB)
Comparing your computed weights to the one used in Matlab’s bkfilter.m, you flipped the order of lower and upper cutoff. It should be:
p_l = 32;
p_h = 6;
