Fed FRB/US, rebuilt
FRB/US
The working paper behind the implementation, with its citation.
An open Python implementation of FRB/US
Working paperNo DOIApril 2026 data vintageVahid Ahmadi · PolicyEngine
What the paper establishes: a from-scratch Python solver for the Board’s published FRB/US matches the Fed’s own pyfrbus on shocked paths at the scale the Fed’s two releases disagree with each other.
The paper’s numbers use the Board’s April 2026 baseline (LONGBASE); the site runs October 2026, so current figures are on the validation tab.
More detailwhat the paper shows, and what moved since April
- The baseline match is an add-factor identity, not evidence.
- The purchases multipliers sit inside CBO’s converted ranges and below the 2010 FRB-US lines in Coenen et al.
- April vintage. The paper, its PDF and its data files (
figures/results.json,figures/pyfrbus_comparison.csv) use the April 2026 LONGBASE. - What moved since. On the October 2026 LONGBASE the purchases multiplier is 0.75 (paper: 0.72) and the 100bp unemployment peak +0.28pp (paper: +0.26pp).
- Expectations. The paper says model-consistent (rational) expectations are not implemented. The model repository now ships them as an opt-in,
Frbus(path, mce=...), gated against the Board’s pyfrbus; VAR expectations stay the default (validation).
Inline PDF viewing is unreliable on mobile browsers, so it’s disabled here. Open the paper in your device’s PDF reader instead.
Open the PDFSource & reproduction: papers/frb-us/ on GitHub · All papers →
Cite this work
Ahmadi, Vahid (2026). “An open Python implementation of FRB/US.” PolicyEngine Macro working paper. https://policyengine-macro.vercel.app/papers/frb-us
@techreport{policyengine2026frbus,
title = {An open Python implementation of {FRB/US}},
author = {Ahmadi, Vahid},
year = {2026},
institution = {PolicyEngine},
url = {https://policyengine-macro.vercel.app/papers/frb-us},
note = {PolicyEngine Macro working paper}
}The underlying model is the Federal Reserve Board’s FRB/US model; please also credit FRB/US as the source model. No DOI is assigned. For an exact version stamp, cite the git commit of the artifact you used from github.com/PolicyEngine/macro.