Have some problems replicating IRFs of financial accels models (loglind). I attach two,
1) Bernanke BGG 1999. I tried to replicate the IRF of figure 5, the main diff is wrt to the 'premium' ( prem=rk-r in the mod).
Besides the magnitudes diff (in y and i), the prem IRF outputs an increase, while fig5 is always negative.
The mod file is taken from an existing dynare post + minor chgs in params.
bggpaper: http://faculty.wcas.northwestern.edu/~lchrist/papers/handbook.pdf
bggmodel:
bggmodel.mod- (1.83 KiB) Downloaded 264 times
2) Hall 2001, based on BGG for UK data. The mod file is as in the paper's appx, ~BGG.
I tried to replicate chart 15 IRFs but got very diff results.
Might be a typo in eqn A12, last term uses 'rk' , which is not defined in the system. Its replaced by '1' to run it.
hallpaper:
- hallpaper.pdf
- (170.85 KiB) Downloaded 253 times
hallmodel:
- hallmodel.mod
- (1.82 KiB) Downloaded 246 times
Any comments/opinions r very wellcome
Cheers,
euclidean