Help: Replicating Del Negro et al (2015) or FRBNY model

[tkrahnke]

I am trying to replicate the DSGE model by Del Negro et al. (2015): "Inflation in the Great Recession and New Keynesian Models" using Dynare. The model is basically a SW07 model with financial frictions and a time-varying Inflation target. The Matlab model-code is published on the website (see also attached the file containing the equilibrium conditions). The model is very similar to the one used (and published) by the Fed New York (https://github.com/FRBNY-DSGE/DSGE-2015-Apr).

Unfortunately, my code is not running because the BK-condition is not satisfied (i.e. I have one forward-looking variable too much). This usually speeks for a timing problem but I cannot find the mistake. I have as many forward-looking variables as the authors have in their Matlab-code (where the model is in Sims form). Some time ago, other Dynare-users published similar questions in this forum (viewtopic.php?f=1&t=6802) but did not post whether they managed to replicate the model in Dynare.

If anybody has already successfully replicated one of the above-mentioned models in Dynare I would be so grateful if she/he would share the respective file/equations with me. Or can somebody give me a hint where a problem in my code could lie?

Thanks a lot in advance and best wishes,
Tobias

[jpfeifer]

One thing the unstable version notes is

ERROR: If the model is declared linear the second derivatives must be equal to zero.
The following equations had non-zero second derivatives:
* Eq # 11

[tkrahnke]

Dear Johannes,

Thank you so much for your hint!!! This brought me to the respective equation and I indeed found a small mistake in that eqaution which solved the entire problem.

Just one last question: How did you get this very detailed ERROR-message (i.e. the one whichn pointed you to look at equation #11)?

Thanks a lot again and best wishes!

Tobias

[jpfeifer]

1. I was using the Dynare unstable version (to be released as Dynare 4.5 hopefully soon)
2. Would you mind providing the working version of the file for other users once your replication is successful?

[tkrahnke]

Thank you, Johannes!

Yes, sure...attached you can find the working version of my file which runs now and also yields reasonable impulse responses (also very similar to the ones of the model by De Graeve (2008): "The external finance premium and the macroeconomy: US post-WWII evidence", which is quite similar to the one by Del Negro et al. (2015)).

Best,
Tobias