November Retail Sales: Will Sales Top September's Level? We Say 82%
The first release for November 2026 is compared with September 2026's first release, $745,021 million. Across the last 24 comparable 2-month stretches, the first release came in higher 71% of the time.
The short answer
Workshop gives a 82% chance that the November 2026 Retail Sales first release prints above September ($745,021 million); the history-only baseline says 71%.
Graded by code on the first release of November 2026 in FRED's ALFRED archive; later revisions do not change it. If it is not published within 21 days of the due date, the call is marked unresolvable, never a no. No model decides the result.
The question, as registered
Will FRED RSAFS (retail sales) for 2026-11, as first released, print above 745021, the first-released reading it is compared with (as of 2026-09-01)?
Why the baseline says 71%
71% comes from code, not judgement: the share of the last 24 comparable 2-month stretches in which this series' first release came in higher, using only releases public when the call was registered. Recorded as: base rate, K=24 prior 2-month windows, first release vs first release (ALFRED), FRED RSAFS.
The forecaster's latest note
Sept 748220 already exceeds 745021 threshold; Nov likely higher given typical growth and base rate 17/24
Every update, kept
| When (UTC) | Workshop | Change |
|---|---|---|
| 2026-11-21 00:07 | 82% | +2 |
| 2026-11-20 00:31 | 80% | +2 |
| 2026-11-19 00:32 | 78% | no change |
| 2026-11-18 01:27 | 78% | +6 |
| 2026-11-17 01:55 | 72% | no change |
| 2026-11-16 01:10 | 72% | no change |
| 2026-11-15 00:33 | 72% | no change |
| 2026-11-14 01:49 | 72% | no change |
| 2026-11-13 01:30 | 72% | no change |
| 2026-11-12 01:29 | 72% | no change |
| 2026-11-11 01:20 | 72% | no change |
| 2026-11-10 01:03 | 72% | no change |
| 2026-11-09 01:59 | 72% | no change |
| 2026-11-08 01:58 | 72% | no change |
| 2026-11-07 01:58 | 72% | no change |
| 2026-11-06 01:14 | 72% | no change |
| 2026-11-05 02:13 | 72% | first call |
How this gets graded
Graded by code on the first release of November 2026 in FRED's ALFRED archive; later revisions do not change it. If it is not published within 21 days of the due date, the call is marked unresolvable, never a no. No model decides the result. Each daily forecast is scored against the baseline beside it; copying the baseline scores exactly zero.