G7 HANK Models Global Macro Model v4 Global Macro Model v5 Payrolls Forecasting

Payrolls Forecasting Lab

Name: RoboMacro
Author: RoboMacro

EXPERIMENT

10 independent models — time-series econometrics through gradient boosting — each forecasting Friday's BLS Employment Situation. Methodology

Oct 2026 NFP

BLS Employment Situation — Friday, November 6, 2026

10 days away

Core Ensemble Forecast — Median of 8 Models

Headline consensus excludes VAR and LightGBM (implementation-corrected but weak on recent prints). Full 10-model median shown below for reference.

Nonfarm Payrolls

76k

80% model range: 50k → 310k

Unemployment Rate

4.2%

80% model range: 4.1% → 5.6%

Part-Time Workers

4641k

80% model range: 940k → 4934k

Avg Hourly Earnings YoY

3.25%

80% model range: 3.06% → 3.54%

NFP model disagreement (σ): ±110k · Core range (80%): 50k – 310k · Full 10-model median: 72k

Monthly Track Record — Forecast vs First Print (Last 4 Releases)

All scoring uses the first BLS print (8:30am headline) — not revised figures. Beat/miss compares who was closer; direction shows whether each forecast was too high or too low vs that first print.

Model implementation corrections (Aug 2026)

Some models were producing invalid forecasts due to implementation errors (not weak signal). Historical rows below for corrected models reflect the old broken pipeline until forecasts are re-run.

VAR — Corrected Aug 2026: VAR was fit on mixed units (NFP MoM in thousands vs ICSA ~200k raw claims), so companions dominated and forecasts were systematically wrong (e.g. −155k vs +57k first print in Jun 2026). Fix: per-column z-scoring before VAR estimation plus ±800k forecast cap. Re-run forecasts after this date reflect the corrected pipeline.
LGBM — Corrected Aug 2026: (1) ADPMNUSNERSA was ingested at person scale (~132.7M) beside PAYEMS in thousands. (2) Regime features used raw claims level differences (~200k) beside MoM NFP. (3) LightGBM lacked StandardScaler and used subsample/colsample on tiny tuned feature sets, producing absurd forecasts (e.g. −1,259k Jun 2026). Fix: ADP unit normalization, StandardScaler on base features, percent-change regime signals, adaptive tree hyperparams for small feature sets, ±800k cap. Historical track-record rows before re-run reflect the old pipeline.

Core ensemble vs street on first print: 3/4 releases we beat consensus

Sep 2026

✓ We beat street

Ensemble 58k off (58k too high)

Street 61k off (61k too high)

Aug 2026

✓ We beat street

Ensemble 96k off (96k too low)

Street 109k off (109k too low)

Jul 2026

✗ Street was closer

Ensemble 107k off (107k too high)

Street 103k off (103k too high)

Jun 2026

✓ We beat street

Ensemble 34k off (34k too high)

Street 53k off (53k too high)

Econometric vs Machine Learning medians (first print): Econ beat street 3/4 · ML beat street 1/4 · Econ closer than ML 3/4

Sep 2026

Econometric closer

Econ median 48k off (48k too high)

ML median 99k off (99k too high)

Aug 2026

ML closer

Econ median 122k off (122k too low)

ML median 87k off (87k too low)

Jul 2026

Econometric closer

Econ median 96k off (96k too high)

ML median 117k off (117k too high)

Jun 2026

Econometric closer

Econ median 15k off (15k too high)

ML median 85k off (85k too high)

Model	Oct 2026 NFP Awaiting first print	Sep 2026 First print +29k	Aug 2026 First print +162k Revised to +133k · not scored	Jul 2026 First print -23k Revised to -10k · not scored	Jun 2026 First print +57k Revised to +31k · not scored
Did we beat street? Core 8-model median vs Dow Jones / FinanceFlowAPI — excludes VAR and LightGBM	Pending	✓ We beat street Ensemble 58k off (58k too high) Street 61k off (61k too high)	✓ We beat street Ensemble 96k off (96k too low) Street 109k off (109k too low)	✗ Street was closer Ensemble 107k off (107k too high) Street 103k off (103k too high)	✓ We beat street Ensemble 34k off (34k too high) Street 53k off (53k too high)
STREETStreet Consensus Economist survey benchmark · first-print miss below	—	Forecast 61k above first print Forecast +90k ↑ First print +29k	Forecast 109k below first print Forecast +53k ↓ First print +162k	Forecast 103k above first print Forecast +80k ↑ First print -23k	Forecast 53k above first print Forecast +110k ↑ First print +57k
CORECore Ensemble (8 models) Excludes VAR + LightGBM · headline median · first-print miss below	Pre-release forecast+76kFirst print not yet released	Forecast 58k above first print Forecast +87k ↑ First print +29k ✓ We beat street Ensemble 58k off (58k too high) Street 61k off (61k too high)	Forecast 96k below first print Forecast +66k ↓ First print +162k ✓ We beat street Ensemble 96k off (96k too low) Street 109k off (109k too low)	Forecast 107k above first print Forecast +84k ↑ First print -23k ✗ Street was closer Ensemble 107k off (107k too high) Street 103k off (103k too high)	Forecast 34k above first print Forecast +91k ↑ First print +57k ✓ We beat street Ensemble 34k off (34k too high) Street 53k off (53k too high)
10Full Ensemble (10 models) Includes VAR + LightGBM · reference only	Pre-release forecast+72kFirst print not yet released	Forecast 58k above first print Forecast +87k ↑ First print +29k	Forecast 96k below first print Forecast +66k ↓ First print +162k	Forecast 107k above first print Forecast +84k ↑ First print -23k	Forecast 34k above first print Forecast +91k ↑ First print +57k
Econometric Models Time-series methods grounded in macroeconomic theory · 5 models
ECONEconometric median (5) Median of econometric models · first-print miss below	Pre-release forecast+53kFirst print not yet released	Forecast 48k above first print Forecast +77k ↑ First print +29k ✓ We beat street Ensemble 48k off (48k too high) Street 61k off (61k too high)	Forecast 122k below first print Forecast +40k ↓ First print +162k ✗ Street was closer Ensemble 122k off (122k too low) Street 109k off (109k too low)	Forecast 96k above first print Forecast +73k ↑ First print -23k ✓ We beat street Ensemble 96k off (96k too high) Street 103k off (103k too high)	Forecast 15k above first print Forecast +72k ↑ First print +57k ✓ We beat street Ensemble 15k off (15k too high) Street 53k off (53k too high)
ARNaive Autoregressive (AR)	Pre-release forecast+72kFirst print not yet released	Forecast 68k above first print Forecast +97k ↑ First print +29k #6	Forecast 82k below first print Forecast +80k ↓ First print +162k #2 Beat street	Forecast 96k above first print Forecast +73k ↑ First print -23k #5 Beat street	Forecast 15k above first print Forecast +72k ↑ First print +57k #2 Beat street
SARIMASARIMAX	Pre-release forecast+41kFirst print not yet released	Forecast 13k above first print Forecast +42k ↑ First print +29k #3 Beat street	Forecast 122k below first print Forecast +40k ↓ First print +162k #5	Forecast 67k above first print Forecast +44k ↑ First print -23k #2 Beat street	Forecast 11k below first print Forecast +46k ↓ First print +57k #1 Beat street
VARVector AutoregressionFIXED AUG 2026	Pre-release forecast-208kFirst print not yet released	Forecast 13k above first print Forecast +42k ↑ First print +29k #2 Beat street	Forecast 404k below first print Forecast -242k ↓ First print +162k #10	Forecast 122k above first print Forecast +99k ↑ First print -23k #7	Forecast 396k below first print Forecast -339k ↓ First print +57k #10
UCUnobserved Components	Pre-release forecast+53kFirst print not yet released	Forecast 48k above first print Forecast +77k ↑ First print +29k #5 Beat street	Forecast 136k below first print Forecast +26k ↓ First print +162k #6	Forecast 83k above first print Forecast +60k ↑ First print -23k #3 Beat street	Forecast 32k above first print Forecast +89k ↑ First print +57k #4 Beat street
MIDASAlmon Distributed Lag	Pre-release forecast+135kFirst print not yet released	Forecast 93k above first print Forecast +122k ↑ First print +29k #7	Forecast 25k below first print Forecast +137k ↓ First print +162k #1 Beat street	Forecast 130k above first print Forecast +107k ↑ First print -23k #8	Forecast 37k above first print Forecast +94k ↑ First print +57k #5 Beat street
Machine Learning Models Data-driven methods with non-linear feature interactions · 5 models
MLML median (5) Median of machine-learning models · first-print miss below	Pre-release forecast+80kFirst print not yet released	Forecast 99k above first print Forecast +128k ↑ First print +29k ✗ Street was closer Ensemble 99k off (99k too high) Street 61k off (61k too high)	Forecast 87k below first print Forecast +75k ↓ First print +162k ✓ We beat street Ensemble 87k off (87k too low) Street 109k off (109k too low)	Forecast 117k above first print Forecast +94k ↑ First print -23k ✗ Street was closer Ensemble 117k off (117k too high) Street 103k off (103k too high)	Forecast 85k above first print Forecast +142k ↑ First print +57k ✗ Street was closer Ensemble 85k off (85k too high) Street 53k off (53k too high)
ENetElasticNet Linear	Pre-release forecast+348kFirst print not yet released	Forecast 136k above first print Forecast +165k ↑ First print +29k #9	Forecast 190k above first print Forecast +352k ↑ First print +162k #8	Forecast 229k above first print Forecast +206k ↑ First print -23k #10	Forecast 64k above first print Forecast +121k ↑ First print +57k #6
RFRandom Forest	Pre-release forecast+80kFirst print not yet released	Forecast 25k above first print Forecast +54k ↑ First print +29k #4 Beat street	Forecast 105k below first print Forecast +57k ↓ First print +162k #4 Beat street	Forecast 83k above first print Forecast +60k ↑ First print -23k #4 Beat street	Forecast 30k above first print Forecast +87k ↑ First print +57k #3 Beat street
XGBXGBoost	Pre-release forecast+71kFirst print not yet released	Forecast 7k above first print Forecast +36k ↑ First print +29k #1 Beat street	Forecast 87k below first print Forecast +75k ↓ First print +162k #3 Beat street	Forecast 117k above first print Forecast +94k ↑ First print -23k #6	Forecast 85k above first print Forecast +142k ↑ First print +57k #7
LGBMLightGBM + RegimeFIXED AUG 2026	Pre-release forecast+58kFirst print not yet released	Forecast 99k above first print Forecast +128k ↑ First print +29k #8	Forecast 217k above first print Forecast +379k ↑ First print +162k #9	Forecast 12k above first print Forecast -11k ↑ First print -23k #1 Beat street	Forecast 388k above first print Forecast +445k ↑ First print +57k #9
MLPNeural Network (MLP)	Pre-release forecast+293kFirst print not yet released	Forecast 211k above first print Forecast +240k ↑ First print +29k #10	Forecast 163k below first print Forecast -1k ↓ First print +162k #7	Forecast 223k above first print Forecast +200k ↑ First print -23k #9	Forecast 176k above first print Forecast +233k ↑ First print +57k #8
All models beating street:		5/10	4/10	5/10	5/10
Econometric models beating street:		3/5	2/5	3/5	4/5
ML models beating street:		2/5	2/5	2/5	1/5

“Off” = absolute miss vs first print (lower is better). Green We beat street = core ensemble was closer than consensus. Street row shows how far consensus missed; core ensemble row shows how far we missed. Models marked FIXED AUG 2026 had implementation errors corrected — see note above.