Simulated Brent averages $73 in 2027, below the futures strip
Where does Brent go if the Gulf reopens when prediction markets and oil executives expect, and demand responds as it did after the war began?
In brief
Where Brent goes to the horizon on market odds for the reopening, a survey of oil executives and demand as the war's first months showed it- Brent averages $73 in 2027 at the median of the paths, $10 below the futures strip.→ Exhibit 1
- The site's producers-and-refills model, fitted to these futures, sits near the strip; the simulation's median sits $10 below it because the reopening comes earlier.→ Exhibit 1
- Each month the strait stays shut adds about $1 to the 2027 average, from $66 with a January reopening.→ Exhibit 3
- No timing shift within a year reaches the strip of $83.75 in this model.→ Exhibit 3
- Consumption fell 2.8% against the pre-war forecast in 2026, against the 6.5% the literature's elasticities predicted.→ Exhibit 6
| What would change this view | Trigger | Today | Decided by |
|---|---|---|---|
| December's front-month average leaves the simulated band | December 2026 average below $72 or above $117 | $96.81 average, 1–13 Nov 2026pending | the month's closes, 31 Dec 2026 |
| February's front-month average leaves the simulated band | February 2027 average below $65 or above $112 | $96.81 average, 1–13 Nov 2026pending | the month's closes, 28 Feb 2027 |
| The EIA's December outlook moves the 2027 surplus | 2027 world surplus below 3.0 or above 5.0 /d | 4.0 mb/d, November outlook of 5 Nov 2026pending | EIA Short-Term Energy Outlook, 8 Dec 2026 |
| Traders reprice the reopening | 'normal by 31 December' odds above 25% or below 5% | 4.5%, 16 Nov 2026crossed | Polymarket, each day's close to 31 Dec 2026 |
The reasoning behind each point
- rangeBrent averages $73 in 2027 at the median of the paths, $10 below the futures strip.
Eight paths in ten average between $63 and $86; the mean is $74. The strip is the front-month equivalent of 13 Nov 2026, $83.75; the forecasters' median is $81.9.
- reconcileThe site's producers-and-refills model, fitted to these futures, sits near the strip; the simulation's median sits $10 below it because the reopening comes earlier.
No timing shift up to twelve months reaches the futures in this simulation. Demand fitted to 2026 lowers the median by $2 against the literature's response, so demand narrows the gap.
- timingEach month the strait stays shut adds about $1 to the 2027 average, from $66 with a January reopening.
Paths with no reopening in 2027 (22.5% of them) average $84 at the median. A US strike on Iran landed by 3 Nov 2026, as Polymarket resolved its ceasefire contracts, so no further strike is drawn.
- needsNo timing shift within a year reaches the strip of $83.75 in this model.
The strait shut all year with no strike gives $84 and $76 in December. Only that case reaches the December 2027 contract of $79.98; the default's December median is $62.
- replayConsumption fell 2.8% against the pre-war forecast in 2026, against the 6.5% the literature's elasticities predicted.
March to September 2026, with the front month 52 log points above the February outlook's path. Elasticities scaled to a fifth, three tenths and four tenths of the literature's centre miss 2026 by 1.5 mb/d against 4.0; they are the default.
Next: Producers and refills to 2029 — Brent to 2029 under producers' choices
The paths' median sits $10 below the futures strip
How wide is the range of monthly Brent prices to the horizon, and where does the futures strip sit in it?The median falls from $92 in the last quarter of 2026 to $63 a year later, against futures of $79.
The paths average $63 to $86 in 2027 between the 10th and 90th percentiles; the futures strip averages $83.75. The paths start in Dec 2026, the month after the latest close.
Exhibit 1 · Monthly Brent and its simulated range, against futures and the EIA's path
About this exhibit
The site's producers-and-refills model is fitted to the same futures, so it sits near them. The simulation's median sits $10 below the strip because the markets and the survey time the reopening earlier than the futures imply. Demand fitted to 2026 lowers it by $2 against the literature's response.
More evidence (1)
Exhibit 2 · Simulated Brent median: the selected setting, a shut strait and deep escalation
About this exhibit
A later reopening raises the year's average about $1 a month
How much does the reopening month move the year's average, and what would the futures need?Paths that reopen in January 2027 average $66, and paths with no reopening that year average $84. They are 22.5% of the paths.
Exhibit 3 · Brent's simulated calendar-year average by the month Hormuz reopens
| Reopens | Weight | 10th | Median | 90th | Dec 2027 median | Paths with a strike |
|---|---|---|---|---|---|---|
| Dec 2026 | 4.5% | $55.4 | $63.6 | $70.5 | $57.3 | 0.0% |
| Jan 2027 | 6.8% | $57.9 | $65.6 | $72.3 | $58.2 | 0.0% |
| Feb 2027 | 6.8% | $59.4 | $66.9 | $72.9 | $58.6 | 0.0% |
| Mar 2027 | 10.2% | $61.0 | $68.3 | $74.6 | $58.8 | 0.0% |
| Apr 2027 | 10.2% | $62.6 | $69.5 | $75.7 | $59.3 | 0.0% |
| May 2027 | 10.2% | $64.3 | $71.0 | $77.6 | $59.5 | 0.0% |
| Jun 2027 | 4.1% | $65.6 | $73.1 | $79.6 | $60.5 | 0.0% |
| Jul 2027 | 4.1% | $66.5 | $73.9 | $81.5 | $59.7 | 0.0% |
| Aug 2027 | 4.1% | $68.3 | $75.6 | $82.9 | $60.1 | 0.0% |
| Sep 2027 | 5.5% | $69.3 | $76.6 | $84.5 | $60.9 | 0.0% |
| Oct 2027 | 5.5% | $70.6 | $78.3 | $86.2 | $61.1 | 0.0% |
| Nov 2027 | 5.5% | $72.4 | $79.5 | $87.7 | $61.7 | 0.0% |
| No reopening in 2027 | 22.5% | $75.5 | $83.5 | $92.2 | $76.0 | 0.0% |
Prediction markets weight the months of 2026, and the Dallas Fed's survey of oil executives weights those of 2027.
About this exhibit
No timing shift within a year reaches the futures; a surplus about 2.3 mb/d smaller does. A strait shut all of 2027 with no strike averages $84.
More evidence (2)
Exhibit 4 · Cumulative chance of a Gulf reopening, by source of the odds
About this exhibit
Exhibit 5 · Brent's simulated calendar-year mean under later reopenings and other inputs, against futures
The reopening shifted later
| Reopening | 2027 mean | Dec 2027 mean | No reopening in 2027 |
|---|---|---|---|
| Survey timing | $73.8 | $63.1 | 22.5% |
| 1 month later | $75.0 | $63.5 | 22.5% |
| 2 months later | $76.2 | $64.6 | 28.0% |
| 3 months later | $77.4 | $65.7 | 33.4% |
| 4 months later | $78.4 | $66.8 | 38.9% |
| 5 months later | $79.4 | $67.7 | 43.0% |
| 6 months later | $80.2 | $68.5 | 47.1% |
| 7 months later | $81.0 | $69.2 | 51.2% |
| 8 months later | $81.8 | $71.0 | 61.4% |
| 9 months later | $82.5 | $72.6 | 71.6% |
| 10 months later | $83.1 | $74.2 | 81.9% |
| 11 months later | $83.4 | $75.4 | 88.7% |
| 12 months later | $83.7 | $76.4 | 95.5% |
| The futures strip | $83.75 | $79.98 | – |
One lever at a time
| Lever | Set to | 2027 mean | Dec 2027 mean |
|---|---|---|---|
| agency surplus overstated (balance error fixed, mb/d) | −0.45 | $73.0 | $62.0 |
| agency surplus overstated (balance error fixed, mb/d) | −1.00 | $75.4 | $65.2 |
| agency surplus overstated (balance error fixed, mb/d) | −1.50 | $77.6 | $68.2 |
| agency surplus overstated (balance error fixed, mb/d) | −2.00 | $80.0 | $71.3 |
| agency surplus overstated (balance error fixed, mb/d) | −2.50 | $82.5 | $74.4 |
| agency surplus overstated (balance error fixed, mb/d) | −3.00 | $85.2 | $77.7 |
| restocking 450 mb a year with the share of surplus in OECD stocks fixed | 0.30 | $75.8 | $66.3 |
| restocking 450 mb a year with the share of surplus in OECD stocks fixed | 0.55 | $71.3 | $61.7 |
| stock–price slope ($ per 100 mb) | 4.00 | $82.0 | $72.6 |
| stock–price slope ($ per 100 mb) | 12.00 | $67.7 | $57.7 |
| OPEC+ holds 1 mb/d below the EIA's path after restoration | −1.00 | $75.2 | $66.6 |
| The futures strip | – | $83.75 | $79.98 |
About this exhibit
The widest input swings the year's mean by $8
Which inputs move the year's mean most, and what did the months since the war began say about demand?Stock–price slope takes the 2027 mean from $78 to $70. Reopening timing swings it by $6.
Exhibit 6 · How far each input moves Brent's simulated calendar-year mean, widest swing first
Each bar holds one input at the low or high end of its draws, with every other input drawn as usual.
About this exhibit
Consumption fell 2.8% against the pre-war forecast in 2026, where the literature's elasticities predicted 6.5%. The default scales them down to fit that year.
More evidence (1)
Exhibit 7 · World oil consumption's shortfall from the pre-war forecast, actual and modelled
The windows: all-in elasticities
| Window | Consumption | Non-Gulf supply | Log price gap | Demand elasticity, all-in | Supply elasticity | Literature model: consumption |
|---|---|---|---|---|---|---|
| 2Q26 | −4.18% | 1.94% | 0.540 | −0.077 | 0.036 | −6.96% |
| 3Q26 | −2.11% | 1.64% | 0.519 | −0.041 | 0.032 | −6.20% |
| Mar–Sep 2026 | −2.83% | 1.77% | 0.524 | −0.054 | 0.034 | −6.47% |
Price per mb/d of Gulf shortfall
| Window | Price gap, $ | Gulf shortfall, mb/d | All-in, $ per mb/d | Fair-value part | Premium part |
|---|---|---|---|---|---|
| 2Q26 | $40.7 | 12.79 | $3.18 | $1.15 | $2.03 |
| 3Q26 | $37.3 | 7.52 | $4.96 | $2.38 | $2.57 |
About this exhibit
The settings put the 2027 median between $73 and $84
How do the timing, demand and strike settings compare?Eight settings cross four reopening timings, two demand responses and one strike risk after December.
Exhibit 8 · Brent's simulated calendar-year average under each setting
| Reopening timing | Demand | Strike risk | 2027 median | 10th–90th | Mean | Dec 2027 median | Above $100 | Strike drawn |
|---|---|---|---|---|---|---|---|---|
| Survey timingdefault | 2026-calibrated | None after the strike | $73.3 | $63–$86 | $74.0 | $61.7 | 0.3% | 0.0% |
| Survey timing | Literature | None after the strike | $75.5 | $67–$85 | $75.9 | $65.3 | 0.1% | 0.0% |
| 3 months later | 2026-calibrated | None after the strike | $77.0 | $67–$88 | $77.4 | $63.7 | 0.4% | 0.0% |
| 3 months later | Literature | None after the strike | $78.3 | $70–$87 | $78.6 | $66.4 | 0.2% | 0.0% |
| 6 months later | 2026-calibrated | None after the strike | $80.0 | $71–$90 | $80.3 | $65.9 | 0.7% | 0.0% |
| 6 months later | Literature | None after the strike | $80.7 | $74–$89 | $80.9 | $67.8 | 0.3% | 0.0% |
| No reopening in 2027 | 2026-calibrated | None after the strike | $83.6 | $76–$92 | $83.8 | $75.8 | 1.3% | 0.0% |
| No reopening in 2027 | Literature | None after the strike | $83.7 | $77–$91 | $84.0 | $75.8 | 0.6% | 0.0% |
With the default's other choices, timing moves the median by $10.3, demand by $2.2 and strike risk by $0.0.
About this exhibit
Each setting is the spread of paths its inputs imply, not a forecast. The Appendix lists every input with its source and date.
Appendix
How one path is made, where each input comes from, and the checks the run passes107 checks, all pass
Brent, month by month from December 2026 to the end of 2027, from a joint model of oil output, consumption and stocks. Each path draws every input once: when the Gulf reopens, timed by prediction markets for 2026 and the Dallas Fed's survey of oil executives for 2027; how much demand gives way to price, as it did in 2026; how fast Gulf output returns once the strait is open; and the agencies' balance with its error. The paths start in the first full month after the latest close, with the strait still restricted then. Every input is read again as the start moves on. The futures strip and the EIA's own path are drawn beside the fan as what the market and the agency say. A setting is a choice the reader makes with the controls, never a figure anyone published. Model-conditional: the ranges are what these inputs imply, not a forecast.
One path, month by month
| Symbol | Meaning | From | Assumed or estimated |
|---|---|---|---|
| the reopening month, drawn from the timing weights; the restart time constant | Polymarket; Dallas Fed survey; Wood Mackenzie | – | |
| demand and supply elasticities at three horizons, the asymmetry, the income elasticity | note A (literature); the 2026 replay | – | |
| fair value of OECD commercial stocks: $76–$84 at end-September 2026, a slope of $4–12 per 100 mb, steeper below normal | Goldman Sachs; ECB; BCCh | – | |
| the premium: $22.45 before news (calibrated), halved at deal news, decaying to a residual once open, lifted in a strike | calibrated; judgement | – | |
| the share of the non-strategic balance reaching OECD stocks; the agencies' balance error | EIA; IEF; the 2026 replay | – | |
| a ceasefire break's Gulf loss and the 30 mb US sale that follows it | Polymarket (timing); Kilian–Plante–Richter; the SPR simulation | – | |
| noise 0.15 while restricted (Kalshi), 0.048 once open with persistence 0.45 (the 2025 residuals); Student t(5) | Kalshi; the 2025 fair-value residuals | – |
Model-conditional: what the inputs imply, path by path; not a forecast. Every elasticity and parameter is a cited range, every event probability a market price or a survey share.
Exhibit 9 · 44 inputs come from publishers, markets and cited research
| Input | In the model | Kind | Source |
|---|---|---|---|
| EIA STEO November 2026: world consumption and production, net stock withdrawals, OECD commercial stocks, the SPR, Brent spot, the Gulf countries, OPEC Middle East capacity and spare | Gulf 26.05 mb/d in November 2026 (its estimate), restored level 29.21; 2027 world surplus 4.03 mb/d; OECD commercial stocks 2,742 mb at end-November | EIA | EIA, Short-Term Energy Outlook, November 2026 (forecast date 5 Nov 2026) — tables 2, 3a, 3b, 3c, 3d, 4a, 5 Nov 2026 |
| EIA STEO February 2026 (5 Feb, the last pre-war edition): the no-war path for the 2026 replay | no-war path: consumption 102.91–105.26 mb/d and Brent spot $55–$62, March–September 2026 | EIA | EIA, Short-Term Energy Outlook, February 2026 — tables 2, 3a, 3b, 3c, 3d, 4a, 5 Feb 2026 |
| EIA STEO January 2026: a cross-check on the February edition's history | its world consumption history differs from February's by at most 0.363 mb/d over January–November 2025 | EIA | EIA, Short-Term Energy Outlook, January 2026 — table 3a, 8 Jan 2026 |
| Brent front month (BZ=F) daily closes: the monthly averages since 2016, the October 2026 level, the 2025 fair-value residuals | November 2026 average $96.81 over 10 sessions to 13 Nov 2026, an estimate until the month ends; monthly means since Jan 2016 | Yahoo Finance | Yahoo Finance, BZ=F daily closes — completed sessions, 13 Nov 2026 |
| Brent delivery-month contracts, December 2026 to February 2028 (the strip), December 2028 and 2029 (context): closes of the last completed session | February 2027 contract $93.03; 2027 strip $83.75 (front-month equivalent) and $86.34 (delivery months); December 2027 $79.98; closes of 13 Nov 2026 | Yahoo Finance | Yahoo Finance, Brent delivery-month contracts (NYMEX BZ) — BZG27.NYM, BZH27.NYM, BZJ27.NYM, BZK27.NYM, BZM27.NYM, BZN27.NYM, BZQ27.NYM, BZU27.NYM, BZV27.NYM, BZX27.NYM, BZZ27.NYM, BZF28.NYM, BZG28.NYM, 13 Nov 2026 |
| Dallas Fed Energy Survey Q3 2026, special question 2: by the end of which quarter crude exports from the Persian Gulf return to normal (107 firms, fielded 16–24 Sep 2026) | 2026Q4 6.54%; 2027Q1 18.69%; 2027Q2 28.04%; 2027Q3 11.21%; 2027Q4 14.95%; 2028+ 20.56% (n = 107, Sept. 16–24, 2026) | Federal Reserve Bank of Dallas | Dallas Fed Energy Survey, third quarter 2026, special question 2 — By the end of which quarter do you expect crude oil exports from the Persian Gulf to return to normal levels?, 30 Sept 2026 |
| Polymarket: “Strait of Hormuz traffic returns to normal by” October 31, November 30, December 31 — the yes price (cumulative odds) | by November 1.0%; by December 4.5%; resolved and left out: 31 October | Market price (yes) | Polymarket, “Strait of Hormuz traffic returns to normal by …?” — the yes price of each dated contract, 16 Nov 2026 |
| Polymarket: “US-Iran ceasefire continues through …?” — the mid of best bid and ask of the November 30 and December 31 contracts | a US strike on Iran landed by 3 Nov 2026, as Polymarket resolved its ceasefire contracts, so no further strike is drawn | Market price (mid of bid and ask) | Polymarket, “US-Iran ceasefire continues through...?” — no contract on the horizon, 16 Nov 2026 |
| Polymarket: “US-Iran Hormuz Agreement by …?” November 30 and December 31 — the deal-news cross-check | by 31 December 14.5% | Market price (yes) | Polymarket, “US-Iran Hormuz Agreement by...?” — December 31 contracts, 16 Nov 2026 |
| Demand elasticity of world consumption to Brent: the literature centre and its 80% range at 1 month, 1 quarter and 12 months | centre −0.06 / −0.08 / −0.11; 80% range −0.030…0.15 to −0.040…0.18 to −0.060…0.22 | Literature | research note: oil demand and supply elasticities (18 Oct 2026) — Recommended parameter distributions, rows 1–3 (Hamilton 2009; Caldara–Cavallo–Iacoviello 2019; Labandeira et al.; Havranek et al.), 18 Oct 2026read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| The 2026 demand calibration: the literature centre scaled ×0.2 at 1 month, ×0.3 at 1 quarter, ×0.4 at 12 months (the lead); ×1 is the literature setting | ×0.2 / ×0.3 / ×0.4 at 1 month, 1 quarter and 12 months (the lead); ×1 is the literature setting | 2026 replay (proposed decision BP2) | research note: the 2027 Brent simulation (18 Oct 2026) — §3a: two scalars fitted to March–September 2026 (impact ×0.1, lags ×0.4, rationing 0.2, 4-month half-life; RMSE 1.22 against 4.05 mb/d); the lead takes ×0.2 / ×0.3 / ×0.4, 18 Oct 2026 |
| Response to price falls relative to rises | 0.5–1.0 | Judgement on Gately–Huntington | Gately & Huntington, The Energy Journal 23(1), 2002 — asymmetric demand responses to price rises and falls, 1 Jan 2002read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| Income elasticity 0.55 (80% range 0.3–0.8), on −0.5% world activity per +10% price at 12 months | 0.55, sd 0.195, clipped 0.1–1.0; feedback −0.05 per unit | Literature | research note: oil demand and supply elasticities (18 Oct 2026) — rows 5 and 18 (Hamilton; Gately–Huntington; Baumeister–Hamilton 2019, Table 1), 18 Oct 2026 |
| Supply elasticity of non-Gulf output: the literature centre and 80% range at 1 month, 1 quarter and 12 months; the 12-month part on a 6–11-month lag | centre 0.01 / 0.02 / 0.05; 80% range 0.003…0.04 to 0.005…0.06 to 0.020…0.10 | Literature | research note: oil demand and supply elasticities (18 Oct 2026) — rows 6–8 (Kilian & Patel, Dallas Fed WP 2625, 2026; Newell & Prest; EIA STEO), 18 Oct 2026 |
| Fair value's slope on OECD commercial stocks at normal stocks: $4–12 per 100 mb, mode $8 | $4–12 per 100 mb, mode $8 | Published | Goldman Sachs, via Dow Jones Newswires (TradingView) — Brent's September risk premium estimated at $22 a barrel; about $8 per 100 mb of OECD stocks, 9 Oct 2026read in research note: oil demand and supply elasticities (18 Oct 2026) row 12 |
| The slope's multiplier at the stress level 2,575–2,675 mb of OECD commercial stocks: 1.3–2.5, mode 2.0 | 1.3–2.5, mode 2.0, at 2,575–2,675 mb | Published (ECB), thresholds mapped to OECD | ECB Economic Bulletin 2/2026, box on non-linearities; Banco Central de Chile IPoM June 2026, Box I.2 — price responses 'nearly double' in extreme inventory states; stress and floor stock thresholds, 30 Jun 2026read in research note: oil demand and supply elasticities (18 Oct 2026) rows 13–14 |
| The slope's multiplier 300 mb above normal stocks: 0.5–1.0 (storage-curve convexity) | 0.5–1.0 at +300 mb | Judgement | Gorton, Hayashi & Rouwenhorst, Review of Finance 2013 — the convex relation of the basis to inventories, 1 Jan 2013read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| Fair value at end-September 2026 OECD commercial stocks (2,683 mb): $76–84 around Goldman's $80 | $76–$84 at 2,683 mb (end-September 2026, the October 2026 estimate) | Published, width a judgement | Goldman Sachs, via Dow Jones Newswires (TradingView) — September fair value about $80, risk premium $22, 9 Oct 2026 |
| The demand-destruction ceiling: $155 (Bernstein) to $160–190 (Citi; Banco Central de Chile) | $155–$190 | Published | Bernstein via StreetInsider (10 Mar 2026); Citi via EnergyNow (27 Apr 2026); Banco Central de Chile IPoM June 2026 — expenditure-share limits, 30 Jun 2026read in research note: oil demand and supply elasticities (18 Oct 2026) |
| Share of the non-strategic surplus reaching OECD commercial stocks: the EIA's own 2027 allocation (0.26) to 2026's realised share net of strategic draws (0.44) | 0.26–0.45; 2026 realised 0.30 | EIA (derived); 2026 realised (checked in the replay) | EIA STEO October 2026, Table 3a; research note: the 2027 Brent simulation (18 Oct 2026) §3b — φ = 0.43 of the March–September non-strategic draw came out of OECD commercial stocks, 18 Oct 2026 |
| Agency 2027 balance error: −2.0 to +0.5 mb/d, mode −0.45 (the IEA's balance 0.9 mb/d tighter than the EIA's; OPEC far tighter) | −2.0 to +0.5 mb/d, mode −0.45 | Judgement on the agencies' gap | International Energy Forum, comparative analysis of the agencies' October 2026 reports — the EIA–IEA–OPEC 2027 balances, 15 Oct 2026read in research note: 2027 oil balances and forecasts (18 Oct 2026) §1 |
| Rationed demand per mb/d of Gulf shortfall beyond the EIA's own: 0.15–0.35 in November 2026, fading with a 4–12-month half-life | 0.15–0.35 per mb/d; half-life 4–12 months | 2Q26 decomposition; half-life a judgement | research note: the 2027 Brent simulation (18 Oct 2026) — §1 (1) rationed demand; §3a; lost Gulf product exports 2.8 mb/d (IEA), 18 Oct 2026 |
| Restart time constant after reopening: 1.8–3.5 months (Wood Mackenzie 70% at 3 months, 90% at 6; the EIA's ramp about 1.8) | 1.8–3.5 months | Published | Wood Mackenzie, how quickly can Gulf oil exports recover — 70% back within three months, 90% within six, 29 May 2026read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| Iran's return on reopening: 1.0–1.5 mb/d above the EIA's restored level | 1.0–1.5 mb/d | Judgement | research note: 2027 oil balances and forecasts (18 Oct 2026) — Iran's exports under restriction and on reopening, 18 Oct 2026 |
| OPEC policy after restoration: a 1 mb/d cut, the EIA's held spare, or producing the restored 2.35 mb/d of spare | −1.0 to +2.35 mb/d, mode 0.0 | EIA Table 3d (the spare); range a judgement | EIA STEO October 2026, Table 3d — OPEC Middle East spare capacity restored to 2.35 mb/d from May 2027, 6 Oct 2026read in research note: 2027 oil balances and forecasts (18 Oct 2026) |
| Saudi bypass while restricted: up to 1.2 mb/d from gas-line conversion, half in January 2027 and all from February | 0.0–1.2 mb/d; half in January 2027, all from February | Published | Bloomberg, via gCaptain — Saudi Aramco's East-West line conversion, 1.2 mb/d, 12 Oct 2026read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| UAE bypass while restricted: 0.3–0.9 mb/d on a second Fujairah line, phased January to July 2027 | 0.3–0.9 mb/d; phased January to July 2027 | Published | Kpler — the UAE's second pipeline to Fujairah, mid-2027, 10 Jul 2026read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| Deal news precedes reopening by 2 or 3 months and halves the excess premium | 2 or 3 months; the excess premium halves | Fits Polymarket's agreement odds (20.0% by 30 Nov, 30.5% by 31 Dec) | Polymarket, “US-Iran Hormuz Agreement by …?” — November 30 and December 31 contracts against the reopening odds, 18 Oct 2026read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| A ceasefire break: Gulf output falls 0.5–7 mb/d for 1–6 months and the premium jumps $5–25 | 0.5–7.0 mb/d for 1–6 months; premium +$5–$25 | May 2026 shut-ins (the ceiling); Kilian–Plante–Richter (the premium) | Kilian, Plante & Richter, Dallas Fed WP 2403 — +15–24% per +20pp odds of a 20% shortfall, 1 Jan 2024read in research note: the 2027 Brent simulation (18 Oct 2026) §1b; research note: oil demand and supply elasticities (18 Oct 2026) row 16 |
| Strategic and Chinese restocking once reopened: 100–450 mb a year at the reference pace, full at $75, zero at $95, 1.5× at $65 | 100–450 mb a year; full pace at $75, none at $95, 1.5× at $65 | Published range; the price rule a judgement | Goldman Sachs (June 2026); IEA Oil Market Report October 2026 via Reuters — more than 1 mb/d of global stockpiling in 2027; about 200 mb of non-US strategic stock to replenish, 14 Oct 2026read in research note: 2027 oil balances and forecasts (18 Oct 2026) |
| Residual premium after reopening: $2–8 | $2–$8 | Judgement (the 2022 peak premium was at most $16) | research note: 2027 oil balances and forecasts (18 Oct 2026) — the 2022 premium, 18 Oct 2026 |
| US SPR: the EIA's path (Table 4a) plus a 30 mb sale over three months after a strike | EIA path −21 mb over December 2026, 103 mb back in 2H27; + 30 mb sale over 3 months after a strike | EIA Table 4a; the SPR simulation's lead setting (GD12) | EIA STEO October 2026, Table 4a; the Reserve drawdown simulation's lead setting — a 30 mb sale after a strike, the US share of the IEA actions of 2011 and 2022, 17 Oct 2026 |
| IEA collective action still to deliver outside the US: 56 mb (Europe 46, Asia 10) over November to January | 56 mb over 3 months | Published | IEA, via Reuters and Polity — 331 mb delivered by end-September of the 400 mb action; the balance outstanding, 14 Oct 2026 |
| Residual noise: monthly σ 0.15 while restricted (fitted to Kalshi's middle strikes), 0.048 once open with AR(1) 0.45 (the 2025–Feb 2026 fair-value residuals); Student t(5) | σ 0.15 restricted, 0.048 open; ρ 0.45; t(5). Residuals Jan 2025–Feb 2026: sd 0.055, lag-1 0.44, innovation sd 0.0495 | Kalshi fit; 2025 residuals (checked in the module) | research note: the 2027 Brent simulation (18 Oct 2026) — §1b residual noise; §4 Kalshi WTI strikes, 18 Oct 2026 |
| EIA Brent spot to the front month: $6 in November 2026 falling linearly to $1 by June 2027, the note's judgement | $6 in November 2026 to $1 by Jun 2027, the note's judgement. November so far: $7.19, EIA's November estimate of $104.00 less the front month's average over 10 sessions to 13 Nov 2026. October $15.59; $0.93 on average in 2025 | The note's judgement, made on 18 Oct 2026 against October's $6.9: EIA's October estimate less the front month's average over 12 sessions to 16 Oct 2026, part of the month | EIA STEO Table 2 against the front month's monthly means — the Dated-type spot against the ICE front month, 16 Oct 2026read in research note: the 2027 Brent simulation (18 Oct 2026) §1b |
| The survey's 'normal by' date read one month earlier as the reopening month, uniform within its quarter | 1 month earlier; the survey's cumulative 'normal by' January 2027 is 12.8% against Polymarket's 4.5% by 31 December | The one shift that matches Polymarket at 31 Dec 2026 (proposed decision BP3) | research note: the 2027 Brent simulation (18 Oct 2026) — §1b reopening month R; §2b, 18 Oct 2026 |
| OPEC other liquids outside the Middle East, taken off OPEC NGLs in the Gulf construction: 0.92 (October edition), 0.93 (February edition) | 0.92 mb/d (October edition), 0.93 (February edition) | The June 2025 – January 2026 gap between the formula and the country rows (checked) | EIA STEO tables 3a, 3c, 3d — the Gulf construction: capacity less spare plus Middle East NGLs, Qatar and Bahrain, 6 Oct 2026 |
| IEA collective action by end-September 2026: 331 mb released, about 2.2 mb/d in March–June and 270 kb/d in September (the replay's non-US releases) | 331 mb by end-September: US SPR 132.3, other members 198.7 | Published; the taper a judgement | IEA via Polity and Reuters — 331 mb of the 400 mb action delivered by end-September, 14 Oct 2026 |
| The no-war price path: the February edition's Brent spot less the pre-war spot-to-front-month spread of about $1 | $1; 2025 mean $0.93 | 2025 monthly spreads (checked) | EIA STEO Table 2 against the front month's monthly means, 2025 — the pre-war spread, 16 Oct 2026 |
| Goldman's 2027 mapping: Gulf output against February 2026 at +1, −0.3 and −5 mb/d gives Brent in the $60s, $80 and about $125 | bull +1 → $60s; base −0.3 → $80; bear −5 → about $125 | Published | Goldman Sachs, via BOE Report/Reuters (11 Oct 2026) and NewKerala (12 Oct 2026) — the 2027 Brent cases by Gulf output, 11 Oct 2026read in research note: 2027 oil balances and forecasts (18 Oct 2026) §1 |
| Forecasters' median for 2027 Brent, the set the note of 18 Oct 2026 collected: $81.9 (EIA 83.74 in its October table, IMF 87.39, Goldman 80, BofA 80, HSBC 85, Rabobank 86, Fitch 70, Citi 66.25) | $81.9 | Derived from the published forecasts | research note: 2027 oil balances and forecasts (18 Oct 2026) — §4a official forecasters; summary statistics, 18 Oct 2026 |
| The EIA's own 2027 Brent spot, the year and 4Q27, from the day's edition | $86.00 for 2027, $77.00 in 4Q27 (spot basis): the mean of the November 2026 workbook's monthly figures, which EIA rounds to the dollar | Published | EIA STEO November 2026, Table 2 (the monthly workbook) — Brent spot, the months of 2027, 5 Nov 2026read in research note: 2027 oil balances and forecasts (18 Oct 2026) §1 |
| Kalshi WTI year-end strikes (touch probabilities), read as a December Brent average by the reflection rule (÷2) and a mid-month factor (×0.9), with the December 2026 Brent–WTI spread of $10.84 | above 115 39.6%; above 120 31.8%; above 125 19.8%; above 150 9.0%; above 200 4.6%; below 65 11.5%; below 60 8.0%; Brent − WTI $10.84 | Market price (daily touch); the reading a judgement | Kalshi, WTI maximum price contracts (KXWTIMAX), about 18 Oct 2026 — settles above 115 / 120 / 125 / 150 / 200 and below 65 / 60 by 31 Dec, 18 Oct 2026read in research note: the 2027 Brent simulation (18 Oct 2026) §4 |
| Goldman's published premia: March 2026 about $35–40; September 2026 just over $20 | March 2026 about $35–40 (fair value in the low $60s against about $100); September 2026 just over $20 | Published | Goldman Sachs, via Reuters (March 2026) and Dow Jones Newswires (9 Oct 2026) — ex-shock fair value and the risk premium, 9 Oct 2026read in research note: oil demand and supply elasticities (18 Oct 2026) §3d |
About this exhibit
Exhibit 10 · The engine reproduces the reference run's 2027 median of $86.77
Model-conditional: every price and range on this page is what the model's inputs imply, drawn from 20,000 seeded paths; it is not a forecast of Brent. The forward-looking probabilities are market prices (Polymarket's reopening and ceasefire contracts), a survey's shares (the Dallas Fed Energy Survey's reopening quarter) and the share of model paths; none is the author's.
- rolling
The paths start in December 2026, the month after that of the latest close, 13 Nov 2026, and run to December 2027. November 2026, the month before the start, enters at the front month's average over its 10 sessions so far, an estimate until the month ends. That average follows the contract trading as the front month each session, the next one when a contract stops a session early at the year's turn. Each month the start moves on, and the inputs are read again from that day's sources. The starting stocks and the Gulf's level are the EIA's for November 2026, from its November 2026 outlook, its estimate for that month. The premium is set again so that December's mean equals the February 2027 contract. The reopening odds are the listed prediction markets' for December, then the Dallas Fed's third quarter 2026 survey, given no reopening by 31 December. A US strike on Iran landed by 3 Nov 2026, as Polymarket resolved its ceasefire contracts, so no further strike is drawn. The paths are conditional on the strait still being restricted at the start of December 2026. November's own market odds are not in the weights or the strike hazard: 'normal by 30 November' at 1.0%. Model-conditional: the start and every input are the day's, and the paths are what they imply, not a forecast.
- changed
Changed by the author's decision of 5 Nov 2026 (D-2026-11-05-BP3R). After the market months the survey is now taken given no reopening by 31 December, in place of the splice used since 18 Oct 2026. On the inputs of 17 Nov 2026 the default's 2027 median is $73.27 against $72.53 under the splice, and its mean $73.95 against $73.34. Paths with no reopening in 2027 are 22.5% against 20.6%.
- changed eia
Changed on 5 Nov 2026: the EIA's own 2027 Brent spot is the mean of the November 2026 workbook's monthly figures, $86.00 for the year and $77.00 for 4Q27. EIA rounds those figures to the dollar. Until then the page printed the annual and quarterly figures of EIA STEO October 2026, Table 2, $83.74 and $73.94, which the workbook does not carry.
- path
Each path runs month by month from December 2026 to December 2027. The Gulf's output starts at the EIA's restricted November level, gains what the Saudi and UAE bypass lines add after it, and restarts toward the EIA's restored level plus Iran and an OPEC policy draw from a reopening month drawn from the timing weights, at a drawn time constant. Deal news arrives two or three months before reopening and halves the excess premium; a ceasefire break, drawn from the market's hazard before any deal news, cuts Gulf output and lifts the premium for a drawn spell and sells 30 mb from the US reserve. Non-Gulf output and consumption respond to the gap between the price and the EIA's own path with the drawn elasticities; part of any Gulf shortfall beyond the EIA's is rationed demand that fades. OECD commercial stocks take a drawn share of the non-strategic balance; the price is the fair value of those stocks plus the premium, times a persistent noise, under a soft ceiling. Price and flows are solved together each month.
- settings
8 settings: the reopening timing as the markets and the survey give it, the survey 3 or 6 months later, or no reopening in 2027; the demand response scaled to 2026 or as the literature has it; a US strike on Iran landed by 3 Nov 2026, as Polymarket resolved its ceasefire contracts, so no further strike is drawn. Each is precomputed; the controls select one. The lead is the survey timing and the 2026-calibrated demand.
- common
One seed (20261018) draws every random input once — the elasticities, the stock and premium parameters, the Gulf's restart and policy, the escalation, the reopening month and the monthly noise — and every setting of one demand response reads the same draws (common random numbers). Two settings therefore differ only by what the setting changes.
- calibration
The premium before news in the restricted state is the one calibrated number: $22.45, set by bisection so that December's mean across the paths equals the February 2027 contract of 13 Nov 2026 ($93.03), and held for every setting.
- timing
For 2026 the reopening odds are Polymarket's: 4.5% by 31 December. For 2027 the survey's quarter shares are spread evenly within each quarter and the 'normal by' date is read one month earlier as the reopening month, the one shift that matches the market at 31 December (12.8% against 4.5%). The survey is taken given no reopening by 31 December, under the author's decision of 5 Nov 2026 (D-2026-11-05-BP3R). Its 12.8% up to then is taken out and the rest scaled to the 95.5% the market leaves. In all, 22.5% of paths see no reopening in 2027. A US strike on Iran landed by 3 Nov 2026, as Polymarket resolved its ceasefire contracts, so no further strike is drawn.
- replay
The 2026 replay runs March to September against the February 2026 outlook, the last before the war: on the realised price the literature's elasticities predicted a consumption fall of 6.5% against the realised 2.8%. The lead scales them to ×0.2, ×0.3 and ×0.4 at 1 month, 1 quarter and 12 months; the literature setting keeps them. OECD commercial stocks took 30% of the non-strategic draw, inside the drawn range.
- defaults
The lead setting and the one-premium calibration are the author's decision of 17 Nov 2026 (D-2026-11-17-BP16). The 2026 demand scale is the author's decision of 5 Nov 2026 (D-2026-11-05-BP2R). The one-month survey shift is part of the author's decision of 5 Nov 2026 (D-2026-11-05-BP3R). The rolling start is the author's decision of 5 Nov 2026 (D-2026-11-05-BP7), and its first month the author's decision of 5 Nov 2026 (D-2026-11-05-BP13).
- precision
Prices are printed to the dollar in the tiles and to the cent in the tables; shares of paths to 0.1 percentage point. A Monte Carlo comparison passes within four standard errors of the difference of two independent runs, from the run's own order statistics for a percentile, the sample standard deviation for a mean and the binomial error for a share, plus half the reference's last printed digit.
- licence
EIA's series carry EIA's attribution; Yahoo Finance's closes are shown with its attribution; the Dallas Fed's shares and Polymarket's prices are quoted with venue and date; the research notes' figures are quoted as published with their documents and dates.
20,000 paths from one seed; the premium before news, $22.45, calibrated once on the default setting so that December's mean equals the February 2027 contract, then held for every setting. November's front month averaged $96.81 over its 10 sessions so far, an estimate until the month ends.
The reference run (the prototype, same inputs and seed) against this run
| Figure | Reference | This run | Within |
|---|---|---|---|
| Lead: 2027 average, mean | 87.26 | 73.95 | 0.35 |
| Lead: 2027 average, median | 86.77 | 73.27 | 0.44 |
| Lead: 2027 average, 10th percentile | 67.17 | 63.11 | 0.43 |
| Lead: 2027 average, 90th percentile | 107.33 | 85.85 | 0.88 |
| Lead: December 2027, median | 63.94 | 61.74 | 0.36 |
| Lead: share of paths with a strike | 0.67 | 0.00 | 0.00 |
| Lead: share reopened by December 2027 | 0.79 | 0.77 | 0.02 |
| 2027 median, strait reopens in Jan 2027 | 70.23 | 65.59 | 0.90 |
| 2027 median, strait reopens in Mar 2027 | 78.75 | 68.27 | – |
| 2027 median, strait reopens after 2027 | 100.20 | 83.52 | – |
| Status quo: 2027 median | 96.61 | 83.56 | 0.36 |
| Status quo: December 2027 median | 91.32 | 75.79 | 0.84 |
| Deep escalation: 2027 median | 123.41 | 106.23 | 0.52 |
| Deep escalation: December 2027 median | 97.51 | 82.50 | 0.78 |
The fair-value curve on 2025: residuals against the front month
| Statistic | $ a barrel |
|---|---|
| Root mean square residual | 3.67 |
| Mean residual | 0.52 |
Goldman's three points, a comparison: the Gulf level of each case run through this model
| Case | Gulf output 2027, mb/d | Goldman, $ | This model, $ |
|---|---|---|---|
| bull +1 | 32.45 | $60s | $49.6 |
| base −0.3 | 31.15 | $80 | $57.4 |
| bear −5 | 26.45 | about $125 | $84.1 |
| bear as status quo (no reopening) | 26.05 | about $125 | $85.6 |
The EIA's own path, a comparison: its November 2026 balance run through this model with no Gulf restriction
| Figure | EIA, futures basis | This model on the EIA balance |
|---|---|---|
| 2027 average | $84.1 | $68.5 |
| December 2027 | $74.0 | $63.8 |
| OECD commercial stocks, December 2027, mb | 3,095 | 2,897 |
| 2027 average, EIA spot: the mean of its monthly figures | $86.0 | – |
Kalshi's year-end WTI strikes, a comparison: the touch odds, the December average they imply, and this model
| WTI strike | Kalshi, touch | Implied December average | Brent equivalent, $ | This model, December average |
|---|---|---|---|---|
| above 115 | 39.6% | 17.8% | $125.84 | 2.4% |
| above 120 | 31.8% | 14.3% | $130.84 | 1.6% |
| above 125 | 19.8% | 8.9% | $135.84 | 1.1% |
| above 150 | 9.0% | 4.1% | $160.84 | 0.2% |
| above 200 | 4.6% | 2.1% | $210.84 | 0.0% |
| below 65 | 11.5% | 5.2% | $75.84 | 8.9% |
| below 60 | 8.0% | 3.6% | $70.84 | 4.0% |
About this exhibit
How the numbers are made
steo_current + steo_feb26 + nonme_ngl ──▶ the EIA's newest outlook read by table: world consumption and production, OECD commercial stocks, the SPR, Brent spot; the Gulf as the country rows in history and as capacity less spare plus Middle East NGLs, Qatar and Bahrain in the forecast; the February edition as the no-war path ──▶ the EIA paths brent_front + brent_contracts + spot_spread ──▶ the front month's monthly averages since 2016 and its level in the month before the start; the strip on the front-month basis — for each month the contract two months out; the EIA's spot path put on the futures basis ──▶ the strip survey_q2 + hormuz_normal + survey_shift ──▶ the reopening weights: Polymarket's cumulative odds for the months its listed 'normal by' contracts price on the horizon; after them the survey's quarter shares, read uniformly within each quarter and one month earlier, given no reopening by the last market month; the remainder is no reopening in the horizon ──▶ the timing weights ceasefire + hormuz_agreement ──▶ the strike hazard from the listed month-end ceasefire contracts on the horizon: one less the first contract's mid for its month, one less each later mid over the one before for its month, none after the last; the agreement contracts as the deal-news cross-check ──▶ the strike hazard demand_elasticity + demand_scale + demand_asymmetry + income_elasticity + supply_elasticity + stock_slope + low_stock_multiplier + high_stock_multiplier + fair_value_anchor + price_ceiling + oecd_share + balance_error + rationing + restart_tau + iran_return + opec_policy + saudi_bypass + uae_bypass + deal_lead + strike_episode + restocking + residual_premium + us_spr_path + other_releases + noise ──▶ the parameter distributions as cited, one draw of each per path from one seed; the open-state noise checked against the 2025 fair-value residuals ──▶ the parameter draws the EIA paths + the strip + the timing weights + the strike hazard + the parameter draws ──▶ themes._brent_mc: the premium before news calibrated so the first month's mean equals its contract; 20,000 paths with common random numbers over the settings; the fixed scenarios on the default's draws ──▶ the paths the paths ──▶ quantiles by month, the 2027 average, the quarters, the paths grouped by reopening month, the sensitivity at the draws' 10th and 90th percentiles, the reconciliation to the strip; seeded rerun, independent seed, mixture and reference checks against the prototype ──▶ the settings the EIA paths + the strip + the parameter draws + iea_action + replay_spread ──▶ the 2026 replay, March to September against the February outlook: quantities on the realised price, the stock share of the draw, the implied premium, the demand fit ──▶ the replay the settings + the replay + goldman_points + eia_brent_2027 + kalshi_strikes + forecasters_median ──▶ the reference table: the prototype's results beside the port's and the published run's; Goldman's points, the EIA's own path and Kalshi's strikes beside the model's; the reproduction checks (validate, check, publish) ──▶ the data the page reads
Sources
- publicU.S. Energy Information Administration: Short-Term Energy Outlook, November 2026 monthly workbook (tables 2, 3a–3d, 4a) and the February and January 2026 archive editions
Source: U.S. Energy Information Administration · https://www.eia.gov/outlooks/steo/xls/STEO_m.xlsx
- licensedYahoo Finance: Brent crude futures (BZ=F) daily closes and the delivery-month contracts December 2026 to December 2029, completed sessions to 13 Nov 2026
Source: Yahoo Finance · https://finance.yahoo.com/quote/BZ%3DF/
- publicFederal Reserve Bank of Dallas: Dallas Fed Energy Survey, third quarter 2026, special question 2 and its chart workbook
Source: Federal Reserve Bank of Dallas · https://www.dallasfed.org/research/surveys/des/2026/2603
- publicPolymarket, Gamma API: “Strait of Hormuz traffic returns to normal by …?”, “US-Iran ceasefire continues through …?” and “US-Iran Hormuz Agreement by …?”, quoted on 16 Nov 2026 at 21:57 UTC
Source: Polymarket · https://gamma-api.polymarket.com/events?slug=strait-of-hormuz-traffic-returns-to-normal-by-december-31
- computed hereFigures quoted from the research notes of 18 October 2026 (elasticities, buffers, balances and published forecasts), each with its document, section and date in the inputs table
Quoted as published; each input names its document and date
- computed hereThe simulation: 20,000 seeded paths with common random numbers over 16 settings, computed here
Model-conditional distributions computed by the site's engine; not forecasts.
Reproduction checks
| Check | Published | Computed | Tolerance | Result |
|---|---|---|---|---|
| Fair value: the anchor's stock level, 2,683 mb, is the end-September 2026 estimate of the October 2026 edition the anchor was cited against (2,683.2 mb in data/raw/eia/2026-10-31; abs, mb) | 2683 | 2683.248864 | 0.5 | pass |
| Fair value: the anchor's edition is October 2026, as the spec cites it (editions that differ) | 0 | 0 | 0 | pass |
| Outlook: the day's workbook is the November 2026 edition, forecast date 5 Nov 2026, on or before 17 Nov 2026 and the newest on disk (newer editions fetched earlier; undated or later) | 0 | 0 | 0 | pass |
| Outlook: the day's edition is at least the latest EIA scheduled on or before 17 Nov 2026 (November 2026, released 10 Nov 2026) (editions behind) | 0 | 0 | 0 | pass |
| Table 3a: world production less consumption equals minus net stock withdrawals in every month of the workbook (72 months; max abs residual, mb/d) | 0 | 0 | 0.000001 | pass |
| Gulf: the forecast construction (OPEC Middle East capacity less spare, plus Middle East NGLs, Qatar, Bahrain and the UAE) reproduces the seven country rows over Jun 2025–Sep 2026 (16 months; max abs difference, mb/d) | 0 | 0.0234 | 0.051 | pass |
| Gulf: the 0.92 mb/d of OPEC NGLs outside the Middle East taken off the construction is within 0.051 of the June 2025 – January 2026 gap the country rows imply (0.934 mb/d) | 0.92 | 0.933525 | 0.051 | pass |
| Grid: the 13 simulated months from Dec 2026 to Dec 2027 all carry the EIA's consumption, production, stocks, spot and SPR (missing values) | 0 | 0 | 0 | pass |
| Outlook: the archive workbook is the February 2026 edition with its forecast date | February 20262026-02-05 | February 20262026-02-05 | 0 | pass |
| Replay: the 0.93 mb/d taken off the February edition's OPEC NGLs is within 0.051 of the gap its country rows imply over June 2025 – January 2026 (0.934 mb/d) | 0.93 | 0.933606 | 0.051 | pass |
| Replay: the February edition's Table 3a identity holds in every month (max abs residual, mb/d) | 0 | 0 | 0.000001 | pass |
| Strip: no simulated month's contract has stopped trading by 13 Nov 2026 (contracts expired) | 0 | 0 | 0 | pass |
| Strip: every live delivery-month contract from BZG27.NYM to BZG28.NYM and the front month have their last completed session on 13 Nov 2026; stopped trading and left out: BZZ26.NYM (contracts on another date) | 0 | 0 | 0 | pass |
| Front month: the BZ=F close of 13 Nov 2026 ($95.84) equals the close of BZF27.NYM, the contract trading as the front month that day (abs difference) | 95.839996 | 95.839996 | 0 | pass |
| Strip: the 2027 front-month-equivalent strip, $83.75, equals its recomputation from the raw files — the twelve closes of BZH27.NYM–BZG28.NYM on 13 Nov 2026 (abs difference) | 83.747499 | 83.747499 | 0 | pass |
| Strip: the December 2027 contract's close ($79.98) is its bar of 13 Nov 2026, the latest completed session, read from the raw file; a bar of an unfinished session after it is not read (abs difference) | 79.980003 | 79.980003 | 0 | pass |
| Strip: the January–December 2027 delivery-month strip, $86.34, equals the mean of the twelve contracts' closes on 13 Nov 2026 read from the raw files (abs difference) | 86.336666 | 86.336666 | 0 | pass |
| Spread: the 2025 mean spread (0.93) rounds to the $1 the no-war path uses (abs difference) | 1 | 0.930566 | 0.5 | pass |
| Survey: the third quarter 2026 survey's special question 2 asks when Persian Gulf crude exports return to normal and carries 6 answers, Q4 2026 to 2028 or later, each a quarter or 'YYYY or later' (answers not read) | 0 | 0 | 0 | pass |
| Survey: the 6 shares sum to 100 (99.99; abs difference, points) | 100 | 99.99 | 0.0101 | pass |
| Survey: 107 firms answered (the workbook's note) | 107 | 107 | 0 | pass |
| Survey: the third quarter 2026 survey's page is dated 30 Sep 2026, the release date the spec lists (30 Sep 2026) | 2026-09-30 | 2026-09-30 | 0 | pass |
| Survey: the 3 shares the survey page prints (Q2 2027 28%, Q1 2027 19%, 2028 or later 21%) match the workbook's rows (28.04, 18.69, 20.56; max abs difference, points) | 0 | 0.44 | 0.5 | pass |
| Reopening odds: the 'normal by' contracts dated on the horizon (31 December) are open, dated by month end and run from its first month without a gap; dated before the start and left out: 31 October (closed), 30 November (open) (closed, missing, undated or after a gap) | 0 | 0 | 0 | pass |
| Reopening odds: no listed 'normal by' contract has resolved that the strait reopened — the restricted start would no longer hold (resolved yes) | 0 | 0 | 0 | pass |
| Reopening odds: the cumulative odds rise with the date — November 1.0%, December 4.5% (falls) | 0 | 0 | 0 | pass |
| Weights: the reopening weights and the 'none' remainder sum to one, and the cumulative odds never fall (abs difference; falls) | 1 | 1 | 0 | pass |
| Weights: December's weight equals Polymarket's 'normal by 31 December' odds (4.5%; abs difference) | 0.045 | 0.045 | 0 | pass |
| Weights: the 'no reopening in 2027' weight (22.52%) is the survey's '2028 or later' share, 20.56%, given no reopening by 31 December, the last market month, scaled to the 95.50% the market leaves, plus the shares' rounding so scaled (22.51%; abs difference, points) | 22.509228 | 22.520177 | 0.011058 | pass |
| Strike: the first ceasefire contract the venue resolved broken ('November 3', ends 2026-11-04) is dated on or before 17 Nov 2026, and every later contract has resolved broken too (10 broken; later contracts resolved held, or not yet settled) | 0 | 0 | 0 | pass |
| Engine: the income feedback (−0.05 per unit of income elasticity) is the cited −0.5% of activity per +10% price | -0.05 | -0.05 | 0 | pass |
| Engine: the slope eases to its high-stock multiplier 300 mb above normal stocks, as cited | 300 | 300 | 0 | pass |
| Engine: the restocking rule (full pace at $75, one less per $20 above it, at most 1.5×) is the cited one | 75201.5 | 75201.5 | 0 | pass |
| Spread: in the prototype's fetch of 18 Oct 2026, EIA's October spot less the front month's October average over 12 sessions to 16 Oct 2026 (6.93) is the $6.9 the prototype printed, to one decimal | 6.9 | 6.93 | 0.051 | pass |
| Restated spread (24 Oct 2026, old): $6.93 over 12 October sessions to 16 Oct 2026, EIA's October figure an estimate, is the strip of 18 Oct 2026, and the entry's source names data/raw/yahoo_brent/2026-10-18 and data/raw/eia/2026-10-18 (spread, sessions, last session, estimate and folders; values that differ) | 0 | 0 | 0 | pass |
| Restated spread (24 Oct 2026, new): $4.77 over 17 October sessions to 23 Oct 2026, EIA's October figure an estimate, is the strip of 24 Oct 2026, and the entry's source names data/raw/yahoo_brent/2026-10-24 and data/raw/eia/2026-10-24 (spread, sessions, last session, estimate and folders; values that differ) | 0 | 0 | 0 | pass |
| Calibration: with the premium before news at $22.45, December's mean across 20,000 paths equals the February 2027 contract ($93.03) within $0.05 | 93.029999 | 93.027044 | 0.05 | pass |
| Calibration: the premium lies inside the bisection's bracket, not on an edge | 0 | pass | ||
| Quantiles: the 5th to 95th percentiles are ordered in every month of every setting (8 settings × 13 months; violations) | 0 | 0 | 0 | pass |
| Seed: a second run of the default setting from seed 20261018 reproduces every path (max abs difference, $) | 0 | 0 | 0 | pass |
| Settings: a later reopening never lowers the 2027 median — survey, 3 months later, 6 months later, none — in any demand and strike setting (falls) | 0 | 0 | 0 | pass |
| Reopening: the share of the default's paths reopened by December 2027 (77.39%) is one less the 'none' weight (77.48%), within 4 binomial standard errors | 0.774798 | 0.77395 | 0.011831 | pass |
| Strike: a US strike on Iran landed by 3 Nov 2026, as Polymarket resolved its ceasefire contracts, so no further strike is drawn; the default draws none (paths with a strike) | 0 | 0 | 0 | pass |
| Independent seed: the default's 2027 average, mean, from seed 20261019 ($73.92) within 4 standard errors of the difference ($) | 73.951634 | 73.922318 | 0.354246 | pass |
| Independent seed: the default's 2027 average, 10th percentile, from seed 20261019 ($63.20) within 4 standard errors of the difference ($) | 63.107536 | 63.198026 | 0.453169 | pass |
| Independent seed: the default's 2027 average, 50th percentile, from seed 20261019 ($73.32) within 4 standard errors of the difference ($) | 73.267959 | 73.317412 | 0.434738 | pass |
| Independent seed: the default's 2027 average, 90th percentile, from seed 20261019 ($85.87) within 4 standard errors of the difference ($) | 85.849066 | 85.874574 | 0.8637 | pass |
| Independent seed: the default's December 2027 median ($61.58) within 4 standard errors ($) | 61.739906 | 61.577597 | 0.459809 | pass |
| Conditional: the groups partition the default's paths and their count-weighted means equal the 2027 mean (abs difference, $) | 73.951634 | 73.951634 | 0 | pass |
| Conditional: every default path is in one reopening group (paths unassigned) | 20000 | 20000 | 0 | pass |
| Conditional: the weight-weighted mixture of the groups' means ($73.91) equals the lead's 2027 mean ($73.95) within 4 standard errors of the weights' sampling | 73.951634 | 73.907699 | 0.251203 | pass |
| Conditional: a path whose strait reopens at or before the shortest deal lead carries no strike (deal news precedes it; strikes in those groups) | 0 | 0 | 0 | pass |
| Scenarios: the status quo draws no strike and the deep escalation a strike on every path (paths off) | 0 | 0 | 0 | pass |
| Scenarios: the deep escalation's 2027 median sits above the status quo's, which sits above the lead's ($, violations) | 0 | 0 | 0 | pass |
| Sensitivity: the base at 8,000 paths ($73.84) is within 4 standard errors of the 20,000-path mean ($73.95) | 73.951634 | 73.844975 | 0.469548 | pass |
| Sensitivity: the widest swing is named from the ranking (stock–price slope, $7.80; the next, reopening timing, $6.45) | 0 | pass | ||
| Reconciliation: the 2027 mean rises with every month of timing shift (falls) | 0 | 0 | 0 | pass |
| Reconciliation: the 2027 mean rises as the fixed balance error falls (violations) | 0 | 0 | 0 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the calibrated premium ($14.11) is the prototype's 14.11 to the cent | 14.11 | 14.11 | 0 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's avg27 mean (87.2633) is the prototype's 87.2633 (abs difference, plus half its last printed digit) | 87.263283 | 87.263283 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's avg27 p50 (86.7650) is the prototype's 86.7650 (abs difference, plus half its last printed digit) | 86.765018 | 86.765018 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's avg27 p10 (67.1654) is the prototype's 67.1654 (abs difference, plus half its last printed digit) | 67.165389 | 67.165389 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's avg27 p90 (107.3271) is the prototype's 107.3271 (abs difference, plus half its last printed digit) | 107.327088 | 107.327088 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's dec27 p50 (63.9386) is the prototype's 63.9400 (abs difference, plus half its last printed digit) | 63.94 | 63.938554 | 0.005001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's p strike (0.67295) is the prototype's 0.67295 | 0.67295 | 0.67295 | 0.00005 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's share open dec27 (0.79185) is the prototype's 0.7919 | 0.7919 | 0.79185 | 0.00005 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's 4Q26 median (106.6442) is the prototype's | 106.64417 | 106.64417 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's 1Q27 median (101.5041) is the prototype's | 101.504109 | 101.504109 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's 2Q27 median (89.7270) is the prototype's | 89.727045 | 89.727045 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's 3Q27 median (79.7040) is the prototype's | 79.704041 | 79.704041 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the lead's 4Q27 median (67.6222) is the prototype's | 67.622166 | 67.622166 | 0.000001 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the 2027 median with the strait reopening in Jan 2027 forced on 10,000 paths (70.23) is the prototype's 70.23 printed to two decimals | 70.23 | 70.233743 | 0.0051 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the 2027 median with the strait reopening in Mar 2027 forced on 10,000 paths (78.75) is the prototype's 78.75 printed to two decimals | 78.75 | 78.751152 | 0.0051 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the 2027 median with the strait reopening in none forced on 10,000 paths (100.20) is the prototype's 100.2 printed to two decimals | 100.2 | 100.195003 | 0.0051 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the status quo's 2027 median on 10,000 paths (96.61) is the prototype's 96.61 printed to two decimals | 96.61 | 96.605073 | 0.0051 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the status quo's December 2027 median on 10,000 paths (91.32) is the prototype's 91.32 printed to two decimals | 91.32 | 91.321066 | 0.0051 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the deep escalation's 2027 median on 10,000 paths (123.41) is the prototype's 123.41 printed to two decimals | 123.41 | 123.409942 | 0.0051 | pass |
| Reference (the prototype's inputs of 18 Oct 2026): the deep escalation's December 2027 median on 10,000 paths (97.51) is the prototype's 97.51 printed to two decimals | 97.51 | 97.507854 | 0.0051 | pass |
| Fair value: the curve at central parameters on the EIA's monthly OECD stocks misses the front month's 2025 monthly means by $3.67 RMSE, below $5 (upper bound) | 0 | 0 | 0 | pass |
| Fair value: the 2025 mean miss ($0.52) is within $2 (abs) | 0 | 0 | 0 | pass |
| Noise: the assumed open-state persistence (ρ = 0.45) is within 0.05 of the lag-1 correlation of the log residuals, January 2025 – February 2026 (0.443; 14 months) | 0.45 | 0.442543 | 0.05 | pass |
| Noise: the assumed open-state σ (0.048) is within 0.005 of the residuals' innovation sd, sd × √(1 − ρ²) = 0.0552 × √(1 − 0.44²) = 0.0495 | 0.048 | 0.049544 | 0.005 | pass |
| Replay: every realised series (consumption, non-Gulf and Gulf output, stocks, the SPR, the front month) is present March–September 2026 (missing) | 0 | 0 | 0 | pass |
| Replay: the strategic draw adds to the IEA's 331 mb (US SPR 132.3 + other members 198.7; abs difference, mb) | 331 | 331 | 0.000001 | pass |
| Replay: 2026's realised φ (0.302) lies inside the drawn range 0.26–0.45 | 0 | pass | ||
| Reference (the prototype's fetch of 18 Oct 2026): the 2025 RMSE equals the prototype's 3.66 printed to two decimals | 3.66 | 3.664941 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the 2025 mean miss equals the prototype's 0.47 printed to two decimals | 0.47 | 0.468332 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the world draw March–September (656.9 mb, Table 3a) equals the prototype's | 656.944813 | 656.944813 | 0.000001 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the share of the non-strategic draw out of OECD commercial stocks (φ = 0.433) equals the prototype's | 0.433186 | 0.433186 | 0.000001 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the literature model's consumption RMSE against 2026 (4.05 mb/d) equals the prototype's 4.05 printed to two decimals | 4.05 | 4.047858 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the two-scalar fit's RMSE (1.22 mb/d) equals the prototype's 1.22 printed to two decimals | 1.22 | 1.223605 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the one-scalar fit's RMSE (1.30 mb/d) equals the prototype's 1.3 printed to two decimals | 1.3 | 1.301588 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the two-scalar fit lands on the prototype's grid point (impact ×0.1, lags ×0.4, rationing 0.2, 4-month half-life) | 0.10.40.24 | 0.10.40.24 | 0 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the replay's monthly table (consumption and non-Gulf output, no-war, realised and literature model; the Gulf; the elastic and rationed parts) reproduces note C's table printed to two decimals (max abs difference, mb/d) | 0 | 0.00497 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the all-in shortfalls by window (consumption and non-Gulf output, realised and literature model, percent) reproduce note C's table printed to two decimals (max abs difference, points) | 0 | 0.00351 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the all-in elasticities and log gaps by window reproduce note C's table printed to three decimals (max abs difference) | 0 | 0.000378 | 0.00051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the dollars per mb/d of Gulf output in 2Q26 and 3Q26 (all-in, fair value and premium parts; the Gulf shortfall) reproduce note C's table printed to two decimals (max abs difference) | 0 | 0.002967 | 0.0051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the 2Q26 and 3Q26 price gaps to the no-war path reproduce note C's table printed to one decimal (max abs difference, $) | 0 | 0.024952 | 0.051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the fair value on realised and no-war stocks and the implied premium, month by month, reproduce note C's table printed to one decimal (max abs difference, $) | 0 | 0.049468 | 0.051 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the replayed OECD stock path (the literature model's balance at the realised φ) reproduces note C's table within 1 mb — the table is printed to the million barrels and the prototype cumulated flows it had printed to two decimals (max abs difference, mb) | 0 | 0.650993 | 1.01 | pass |
| Reference (the prototype's fetch of 18 Oct 2026): the replay's realised monthly balance and strategic flows reproduce note C's table printed to two decimals (max abs difference, mb/d) | 0 | 0.004777 | 0.0051 | pass |
| Range kinds: the fan and the 2027 average's spread are simulated ranges with their quantiles (problems) | 0 | 0 | 0 | pass |
| Reopening odds: the markets' and the survey's months and no reopening by Dec 2027 sum to one (abs difference) | 1 | 1 | 0 | pass |
| Context contracts (OB6): 4 of 4 requested on this day's fetch and on disk (BZH28.NYM, BZM28.NYM, BZM29.NYM, BZZ30.NYM); 0 not requested that day (requested and missing) | 0 | 0 | 0 | pass |
| Grammar: the headline, every tile, takeaway, title and falsifier row keep the house grammar (faults) | 0 | 0 | 0 | pass |
| Grammar: every reader line reads in one pass — subject first, one qualifier, no method word, no sentence over 28 words (86 strings; faults) | 0 | 0 | 0 | pass |
| Decisions: every desk decision this page's specification names is in the decisions file (D-2026-10-24-TH1, D-2026-11-05-OB6, D-2026-11-05-OB8, D-2026-11-05-OB9, D-2026-11-05-BP7, D-2026-11-05-BP1R, D-2026-11-05-BP2R, D-2026-11-05-BP3R, D-2026-11-05-BP4R, D-2026-11-05-BP5R, D-2026-11-05-BP6R, D-2026-11-05-BP8, D-2026-11-05-BP9, D-2026-11-05-BP10, D-2026-11-05-BP11, D-2026-11-05-BP12, D-2026-11-05-BP13, D-2026-11-05-BP14, D-2026-11-05-BP15; ids missing) | 0 | 0 | 0 | pass |
- U.S. Energy Information Administration: Short-Term Energy Outlook, November 2026 monthly workbook (tables 2, 3a–3d, 4a) and the February and January 2026 archive editions — Source: U.S. Energy Information Administration
- Yahoo Finance: Brent crude futures (BZ=F) daily closes and the delivery-month contracts December 2026 to December 2029, completed sessions to 13 Nov 2026 — Source: Yahoo Finance
- Federal Reserve Bank of Dallas: Dallas Fed Energy Survey, third quarter 2026, special question 2 and its chart workbook — Source: Federal Reserve Bank of Dallas
- Polymarket, Gamma API: “Strait of Hormuz traffic returns to normal by …?”, “US-Iran ceasefire continues through …?” and “US-Iran Hormuz Agreement by …?”, quoted on 16 Nov 2026 at 21:57 UTC — Source: Polymarket
- Figures quoted from the research notes of 18 October 2026 (elasticities, buffers, balances and published forecasts), each with its document, section and date in the inputs table — Quoted as published; each input names its document and date
- The simulation: 20,000 seeded paths with common random numbers over 16 settings, computed here — Model-conditional distributions computed by the site's engine; not forecasts.