Brent Forecast This Week
As of October 22, 2026, our registered call for week 43 is that December Brent settles inside $104-114 every session through the end of the week, starting from about $113.19. We put 60% on it before the window opened. Last week 3 of 5 graded calls were right. The running record is 75 right, 34 wrong, every call graded in public and never edited.
YOUR COSTS This is hitting fuel budgets right now. See what it's costing your business, in dollars, per month, and the price move it takes to hold your margin. Open the Fuels desk →This week's board · 6 open calls
| Call | Type | Registered odds | Status |
|---|---|---|---|
| The first Kpler weekly Hormuz crude-flow figure that Reuters or Bloomberg reports for a 7-day week ending Oct 21-25 is between 5.0 and 12.0 mb/d inclusive, AND no new UKMTO/JMIC/CENTCOM halt, mine-strike or mine-find notice Oct 20-23, AND no reported Kpler zero day Oct 20-23 | kinetic-geo | 75% | open |
| IF no verified Hormuz halt occurs through Fri Oct 23, THEN December Brent settles inside $104-114 every session | commodity | 60% | open |
| December gold settles above $4,075 every session through Fri Oct 23 | macro-data | 70% | open |
| November Henry Hub natgas settles below $3.40 every session through Fri Oct 23 | commodity | 90% | open |
| The distillate crack (November ULSD settle x 42 minus December Brent settle, same day) settles below $107 every session through Fri Oct 23 | freight | 70% | open |
| December minus February Brent (same-day settles) holds above $4.50 on every settle through Fri Oct 23 | curve | 80% | open |
Every call resolves on named exchange settles or a named public count, inside a window stated in advance. The full text of each call, with its anchor and its bar, is in this week's briefing.
Last week's board · 3 of 5 right
| Call | Was | Grade |
|---|---|---|
| The first Kpler weekly Hormuz crude-flow figure that Reuters or Bloomberg reports for a 7-day week ending Oct 14-18 is between 5.0 and 12.0 mb/d inclusive, AND no new UKMTO/JMIC/CENTCOM halt, mine-strike or mine-find notice Oct 13-16, AND no reported Kpler zero day Oct 13-16 | 75% | unresolved |
| IF no verified Hormuz halt occurs through Fri Oct 16, THEN December Brent settles inside $102-112 every session | 55% | right |
| December gold settles above $4,110 every session through Fri Oct 16 | 65% | wrong |
| November Henry Hub natgas settles below $3.45 every session through Fri Oct 16 | 90% | right |
| The distillate crack (November ULSD settle x 42 minus December Brent settle, same day) settles below $100 every session through Fri Oct 16 | 65% | wrong |
| December minus February Brent (same-day settles) holds above $4.50 on every settle through Fri Oct 16 | 70% | right |
Does the record show skill?
| Registered odds | Calls | Hit rate |
|---|---|---|
| 50-60s% | 17 | 82% |
| 65-75% | 56 | 80% |
| 80-95% | 8 | 88% |
A calibrated forecaster's 70% calls come true about 70% of the time. Ours have run hotter than their odds, which means the odds were too cautious, not that the calls were wrong; the registered figure now sits within ten points of the model that prices each bar. The week-by-week strip, with the losing runs left visible, is on the track record.
Questions
What is Crude Signal's Brent forecast this week?
For week 43: IF no verified Hormuz halt occurs through Fri Oct 23, THEN December Brent settles inside $104-114 every session. We registered it at 60% before the first settle of the window and grade it on the settles, not on our reading of them.
How accurate have the forecasts been?
Across every resolved call since the record began: 75 right, 34 wrong. Last week 3 of 5 graded calls were right. Calls we registered in the 80-95% range have hit 88% of the time (8 calls); calls in the 50-60s have hit 82% (17 calls). The full record, misses included, is at /track-record.
How is a forecast graded?
Each call names its bar, its window, and the price series that resolves it before the window opens. A settle through the bar on any day grades it wrong; a conditional call whose condition never fires grades void. Grades are never edited after publication.
Is this investment advice?
No. These are falsifiable statements about where contracts will settle, published so the record can be checked. They are not a recommendation to buy or sell anything.