NY Fed updates DSGE forecast
The New York Fed published its March DSGE model update, supplying fresh macro forecasts that fold in global shocks, policy rates, and output gaps for 2026 scenarios. The release provides a clean baseline for scenario analysis and for replicating DSGE-based impulse responses in Python or R. (libertystreeteconomics.newyorkfed.org)
Liberty Street Economics published the March 20, 2026 DSGE update by Marco Del Negro, Ibrahima Diagne, Keshav Dogra, Elena Elbarmi, Donggyu Lee, and Michael Pham. (libertystreeteconomics.newyorkfed.org) The update raises the model’s 2026 real GDP growth forecast to 1.0 percent from 0.6 percent in the December run. (libertystreeteconomics.newyorkfed.org) GDP growth forecasts for 2027–2029 were lowered in the March update to 0.2 percent, 0.9 percent, and 1.3 percent, respectively, down from December’s 0.8 percent, 1.3 percent, and 1.8 percent. (libertystreeteconomics.newyorkfed.org) The authors attribute the 2025–early‑2026 upside surprises mainly to MEI (marginal efficiency of investment) shocks, which the post says likely capture a surge in AI‑related investment in late 2025 and early 2026. (libertystreeteconomics.newyorkfed.org) Inflation is described as more persistent in the March run, with the model pointing to cost‑push shocks—potentially reflecting tariff effects—as key drivers, while projections for the short‑run real natural rate r* remain unchanged from December. (libertystreeteconomics.newyorkfed.org) The forecast uses data through 2025:Q4 augmented with March SPF median nowcasts for real GDP and core PCE and with 10‑year Treasury and Baa corporate yields averaged through February 25; expected federal funds rates one to six quarters ahead were constrained to January Survey of Market Expectations medians. (libertystreeteconomics.newyorkfed.org) The post explicitly notes the DSGE model forecast is a Research staff input—not an official New York Fed forecast—and that the projections were produced before the start of the Iran war and therefore exclude its economic impact. (libertystreeteconomics.newyorkfed.org) Reproducibility tools are available: the FRBNY’s DSGE.jl GitHub repository implements the New York Fed DSGE model and mirrors the MATLAB code referenced in the Liberty Street post, with documentation hosted on the DSGE.jl site. (github.com)
