IRFs on the Fed's EDO Model

Hi everyone,

I’m trying to run the Fed’s EDO model using the replication package available on their website (here). However, the impulse responses I obtain are different from those published in the documentation (Figure 2 here). In particular, real GDP is estimated to be about 0.9% lower at the peak, whereas I estimate a peak drag of 1.8% for a 1pp policy rate shock (the path looks the same,
just twice as large as what they show). My policy rate impulse looks the same as theirs, so I don’t think this is a matter of units on the interest rate.

Has anyone here tried to replicate their IRFs? I guess the calibration they published could be different from that of the documentation, but I’m not sure. They could also be showing the impact of a 1pp increase in the annualized policy rate, but that should imply the shock is smaller by a factor of 4 (since this is quarterly), not 2.

Thank you!

linearized.mod (19.2 KB)