PolicyEngine Macro

model 03 — Bank of England structural VAR · boe-svar · UK · hosted

Explain UK growth and inflation.

Decompose UK GDP and inflation into six structural shocks, forecast with credible bands, and explain revisions between quarters.

with the microsimulation

Its forecast prices CPI-linked benefits through the microsimulation.

→Connection. Forecast-to-incidence bridge: svar_inflation_incidence passes the gap between this model's CPI path and the official one through the April 2028 uprating of CPI-linked benefits in PolicyEngine. Experimental, and labelled as such. No reform scoring — score_reform does not accept this model.
→Worked example. 2027 median CPI runs 0.94pp above the March 2026 EFO; passed through Universal Credit and Child Benefit uprating, that costs £0.09bn in 2028, with the largest relative gains in the second-to-fifth income deciles.
→Validation. Replicates Brignone & Piffer (2025), CI-gated. But the 49-origin rolling evaluation is not a credential: against a random walk with drift, and adjusting the 64 variable×horizon tests together, no variable shows skill at any horizon (minimum q = 0.36). Bank Rate comes closest and does not survive. See the evaluation; the scored record is on forecasts.
→Results. The live UK fan chart on the home page and every archived round on forecasts.
01 — what it is

The Bank's own empirical lens, replicated in the open.

In July 2025 the Bank of England published Macro Technical Paper No. 3 — A structural VAR model for the UK economy — with no replication package. This project re-implements the full pipeline in Python — public-source data assembly, Bayesian estimation, identification, and every output figure — validated against the paper's published results.

Within the suite it is the empirical, time-series member — a structural VAR: a statistical model of a small set of series, with just enough economic structure to name the shocks driving them — and the baseline/conditioning member, saying what is hitting the economy and how uncertain the baseline is before you simulate changing it with the structural members.

Commands live on the code page, estimation and identification choices on the methodology page, and replication checks, forecast scores, and limits on the validation page.

Current boe-svar UK real GDP growth forecast Fan chart from the latest complete quarterly data edge, 2026Q1, forecasting 2026Q2 to 2029Q2. Median with 68 and 90 percent predictive bands. Current forecast from 2026Q1 · year-on-year % +0% +1% +2% +3% +4% 90% / 68% bands · median 2026Q2 2027Q2 2028Q2 2029Q2
Bands are the model’s raw posterior quantiles, not rescaled to measured coverage — the calibrated version is on the forecast record. Live-facing forecast, distinct from the frozen 2024Q2 validation experiment. Data through 2026Q1; forecast begins 2026Q2. Estimated on 1992Q1–2025Q1; 5,600 posterior draws, 385 accepted; five stochastic paths per accepted draw. Source and provenance: papers/boe-svar/figures/current_forecast.json.
02 — what it can answer

What's happening now, and why.

Four tools, all posterior distributions rather than point estimates:

The SVAR toolkit and what each output tells you
toolwhat it tells you
IRFs & FEVDs How each named shock propagates through the UK economy (the impulse response function, IRF), and how much of GDP and inflation variance each explains at every horizon (the FEVD).
Historical decompositions Which shocks drove each episode — how much of the 2022 inflation surge was world energy versus UK demand.
Fan-chart forecasts Unconditional forecasts with 68% and 90% credible bands, plus the probability that each shock hitting the latest quarter was positive or negative.
Forecast-revision decomposition The paper's Section 5 exercise: why did the forecast change since last quarter? Splits the revision into news (new shocks) and reassessment via composite impulse responses.

The current 2026Q1 round reads the latest quarter as most likely hit by negative world-demand (probability 0.71) and world-energy (0.90) shocks, against positive UK demand (0.81) and expansionary monetary policy (0.84). Source and provenance: latest_shocks.json, 6,000 draws, 430 accepted, ESS 198.

It does not score reforms. Mapping a policy reform into identified structural innovations is conceptually dubious, not merely unbuilt. For reform scoring use the OBR emulator (short-run, fiscal) or the OLG model (long-run, structural) — both fed by the PolicyEngine microsimulation's static costing, as the score a reform workflow shows; this model tells you what you are scoring against.

Its forecast does feed the microsimulation the other way. UK benefit uprating is CPI-linked, so the svar_inflation_incidence tool prices the gap between this model's CPI path and the official one: for 2027 the median runs 0.94pp above the March 2026 EFO, and passing that through the April 2028 uprating of a curated slice of CPI-linked benefits (Universal Credit standard allowances and child element, Child Benefit) costs £0.09bn in 2028, with the largest relative gains in the second-to-fifth income deciles. A curated slice, not the full fiscal sensitivity — the state pension and frozen tax thresholds are deliberately excluded — and an experimental illustration built on the forecast median, not a scored reform.

03 — sources

The paper, the code, the method.

Where the SVAR's inputs and outputs live
whatwhere
The source paper Macro Technical Paper No. 3 — Brignone & Piffer, Bank of England, July 2025.
Companion paper Brignone & Piffer, structural forecast analysis — the Section 5 forecast-revision method, later BoE Staff Working Paper No. 1,165 (January 2026).
Our working paper papers/boe-svar — the full replication write-up, quarterly forecast scorecard, and enumerated deviations.
This replication github.com/PolicyEngine/boe-var-model — code, prepared data, figures, and the annotated RESOURCES.md.
Method papers Arias, Rubio-Ramírez & Waggoner (2018) — zero/sign identification · Chan, Matthes & Yu (2025) — permutation-search speedup · Giannone, Lenza & Primiceri (2015) — the prior framework.
Reference toolboxes bsvarSIGNs (R, the exact Arias et al. machinery) · ECB BEAR · Cesa-Bianchi VAR-Toolbox — used in the independent econometrics review, which caught and fixed a genuine bias in the zero-restriction importance weights.