Oil is the lifeblood of the world economy. We refine it into the fuels that grow our food and move our goods. It's in the gas in your car and the underwear you put on this morning.
Since the US and Israel went to war with Iran in February 2026, the world has been using more oil than it produces. The Strait of Hormuz and the Red Sea are effectively closed to oil shipping, cutting off roughly 20% of the world's oil supply. Russian refinery strikes and export bans are making the situation worse. The US Strategic Petroleum Reserve is at its lowest level since 1982.
This site tracks how much buffer is left. It's a scenario model, not a prediction, with three possible futures. The odds change only when something verifiable changes, and every forecast gets scored against what actually happens.
Last updated Sep 12, 2026
Brent
$104.61
Sep 11 settle −2.8% · +38% vs pre-crisis ~$76
US diesel (AAA)
$6.16
Sep 12 · new all-time record ($6.1602) — third straight · +66% vs pre-war $3.72
US gasoline (AAA)
$4.31
Sep 12 · +16¢ in a week (AAA) · +53% vs Jan $2.81
SPR
286.6M
w/e Aug 28 · −128.8M since pre-war 415.4M · lowest since Dec 1982
US diesel & heating oil
104.2M
as of Aug 28 · down 14% vs 5-yr avg · East Coast stocks 27% below last year — the tightest link
Everything money-related in this crisis, in one place. Brent is the world benchmark for crude oil — when the news says "oil price," it means Brent. The gasoline and diesel numbers are US national weekly averages; there is no verified weekly world series for those yet, so the world signal comes through Brent and the two gas benchmarks further down this section. All points are sourced — none are interpolated.
EIA weekly, national
From $2.81 in January to $4.31 now — +53%, with the May peak at $4.50. Every month of 2026 since March has averaged above pre-crisis levels.
AAA national, sourced points
From $3.72 in the last pre-war week (Feb 27) to a new all-time record $6.16 on Sep 12 — a third straight record, after passing $6.00 on Sep 11 — +66%, the same baseline as the stat card above. The old record (June 2022, ≈$5.85) fell on Sep 4. The chart plots weekly anchors — the daily Sep 3–12 readings are in the crack spread chart below. Diesel leads gasoline: product, not crude, is the scarce thing.
TTF, $/MMBtu · filled: EIA weekly avgs (Jan–Apr) · hollow: assessed spot (May–Aug)
TTF is Europe's benchmark gas price — where most of the continent's gas trades. It's up about 50% from before the closure, after hitting a three-year high in March and easing on the June ceasefire. Europe entered winter with storage at 48% against a five-year average of 63%, so it had to buy spot LNG in the exact season the strait closed.
JKM, $/MMBtu · filled: EIA weekly avgs (Jan–Apr) · hollow: assessed spot (May–Aug)
JKM is Asia's benchmark LNG price. It's more than double pre-closure, and set a record — $23.17 — on Aug 19. Qatar's Ras Laffan feeds most of Asia's LNG, and its 3–5 year outage has pushed Asian buyers into global spot markets all summer. Asia stores almost no gas, so JKM tracks spot supply and weather.
Brent: sourced points (EIA monthly avgs: Mar $103.0 · Apr $117.29 · Jul $83.76) · WTI: weekly (FRED)
Observed points only — hollow markers are a source's rounded estimate, not a measured close. The path: $76 before the closure → $126 intraday peak (March, +65% — the largest monthly rise in oil-market history) → $95 by mid-April on ceasefire hopes → $74 on Jun 30 → back through $100 on Sep 9 as the tanker war took hold — then +$6.82 in a single day to $108.03 on Sep 10, then −2.8% to $104.61 on Sep 11. The gray weekly line is WTI (FRED spot; the Sep 10–11 points are the front-month close and the settle) — the benchmark US products price against; it reconciles within about a dollar of news-reported settlements, while FRED's Brent series does not, so it is the crude basis for the crack chart beside it. Physical benchmarks ran higher still: the IEA's North Sea Dated printed $113.48 on Sep 9. And positioning is one-sided — hedge funds and money managers added 12% to their combined Brent/WTI net longs across August (OPEC).
Retail price less WTI, $/bbl · weekly (EIA/AAA − FRED).
The crack is what's left after you subtract crude from the product price — the money for refining, hauling, and retailing. Diesel's has nearly doubled since January: ~$89 to $158.7 on Sep 12 — a third straight all-time diesel reading ($6.16) against WTI, which settled 2.4% lower at $100.05. The path wasn't monotonic: $153.6 on Sep 6, when retail's record streak was running ahead of crude; $148 on Sep 10, the crack compressing — crude jumped faster than the pump. Gasoline's has widened too (~$59 to $78–84), after dipping to $52 in March when crude spiked faster than the pump price. Crude is up 38–45% vs pre-war; the product is up more. At wholesale the squeeze is wider still: the IEA puts US diesel/gasoil above $200/bbl in early September — 94% above pre-war — with the Gulf's and Russia's combined diesel exports down 1.6 mb/d from February. Basis: retail prices (EIA/AAA) less FRED's weekly WTI (the Sep 10–11 points are the front-month close and the settle) — the official EIA crack uses wholesale prices and sits lower, but moves the same way. Six product dates have no same-day FRED WTI value (four Mondays, a Sunday, a Saturday); those cracks use the nearest trading day's WTI (±1–2 days) — no interpolation.
What got lost, drained, or stopped: the world's oil balance flipped from a surplus to a drawdown the month the war began; the bill so far, paid in reserves; the US stockpile draining toward its floors.
The world's oil balance flipped from a ≈4M b/d surplus to net withdrawals the month the war began. Every month since has been paid for in reserves, by the whole world. This week's IEA report puts the full-year supply loss at 5.7 mb/d, about 6% of the world's oil, and sees no full recovery in Middle East flows until 2027. In August, global production fell to 100.1 mb/d, with more than 10 mb/d of Gulf output still shut in. Saudi output alone fell 2.3 mb/d in a single month, to 5.97.
World oil balance, million b/d — production minus consumption, EIA STEO Table 3a, reported months (Jan–Aug actuals; the EIA's forecast tail is not shown) · the physical loss peaked at 11.2M b/d of Gulf shut-in in May — demand destruction and non-Gulf supply absorbed most of it · the IEA's observed-inventories count: 507 mb drawn since February — 2.8 mb/d on average, 95 mb of it in August alone.
What the deficit has consumed so far — every figure verified as of Sep 11.
Global commercial stocks
−400M bbl
year-to-date · EIA est. (Sep 9)
US Strategic Petroleum Reserve
−129M bbl
since Feb 28 · EIA
IEA coordinated release
400M bbl
pulled from 32 countries · IEA
China commercial stockpiles
~2–3M b/d
draw inferred from customs data · official SPR untouched
Demand destruction
−2.5M b/d
full-year 2026, cut from −1.6 in the August edition · IEA OMR, Sep 11
Still in the hole
−1.8M b/d
Q4 2026 forecast — supply still below demand · IEA OMR, Aug 12
EIA weekly ending stocks, million bbl
The Strategic Petroleum Reserve (SPR) is the US government's emergency stockpile of crude oil, built after the 1970s energy shocks. It held 415.4M barrels when the war began; it holds 286.6M now. It is being drawn to keep fuel flowing while Gulf barrels are stranded, and its pace is the clearest gauge of how badly the market is hurting.
The red lines are floors, not forecasts: ~300M is the cavern-damage threshold — below it the caverns cannot safely be refilled after a draw (breached Aug 14). 250M is the GEF operational minimum for a sustained draw, 180M the hard operable limit, and 70M the DOE's stated safe minimum — the draw stops there, it does not continue. The dashed paths project three draw rates from the Aug 28 level out to ~March 2027: 0.45M b/d (corridor holds, ~205M by then), 0.70M b/d (standoff, reaches 250M in late October and ~160M by March), 1.35M b/d (corridor lapses, through 250M within a month and pinned at the 70M floor from ~Feb 2027). The EIA's report lands every Wednesday for the week ending 12 days earlier — Wednesday's print (Sep 9) was the Aug 28 week, so the next point on the chart arrives Sep 16 (the Sep 4 week).
Probability of each track — reweighted only on regime events.
regime event: ESPO suspension (the bypass is offline) + Houthi hold of the entire Red Sea coast — reverts to 10/50/40 on an official restart
regime event: 10 tankers/week + $100 + Jazan
regime event: exclusion zone + first third-country base hit
post-8.6M-claim bust
original model (de-escalation 65%) — approx
Corridor holds
10%
Transit resumes under some regime — an Iran–Oman safe corridor or US-escorted lanes — and Hormuz trends back toward normal over one to two quarters.
Brent mean-reverts toward $70–80; the SPR draw slows to ~0.45M b/d and the depletion timeline stretches out again.
Standoff drift
40%
The war continues at current intensity — periodic tanker strikes, the exclusion zone holding, Iranian infrastructure partially offline, the Saudi bypass pipeline struck, no restart date — but no sustained closure.
Brent ranges $95–125; the SPR draws ~0.70M b/d; global stocks keep falling; the breaking-points cascade is delayed, not cancelled.
Corridor lapses
50%
Sustained closure or escalation — tanker losses spike, the exclusion zone hardens into permanence, the bypass stays shut, Abqaiq and Jazan stay offline for months.
Brent $130+; the SPR drains at 1.35M b/d and the inventory cascade begins — US East Coast first, then Russia, Europe, China, and jet fuel.
The odds are judgment weights that move only on pre-decided events; a quiet week is not evidence. The full rule, the reversion clause, and the dated triggers live onthe model page.
The stockpile's whole history on one scale: 727M at the December 2009 peak, 294M at the previous record low in December 1982, and 286.6 now. The green dash marks where the war began. At the 70M floor the draw stops — it does not continue.
The release began in earnest Apr 3 (413.3M) and ran fastest in May (≈1.2M b/d). The draw has slowed to ≈0.45M b/d — the corridor-holds pace — while diesel and heating-oil stocks sit 14% below their 5-year average.
SPR
286.6M bbl
−128.8M (−31%) since pre-war 415.4M
lowest since Dec 1982
US diesel & heating oil
104.2M bbl
down 14% from the 5-year average (121.2M bbl)
East Coast stocks 27% below last year — the tightest link in the chain
US crude
424.5M bbl
+1% vs 5-yr avg (420.3M)
products, not crude, are the scarce thing
Global inventories
−400M bbl YTD
EIA estimate, Sep 9
falling through end of 2026
How fast the stockpile is draining, and when the path hits the floors — EIA weekly, as of Sep 9.
SPR drawdown rate
0.65million barrels/day
4-week average, EIA weekly report (week ending Aug 28) · latest single week: 0.45.
Weekly drain in millions of barrels — May peak of 9.9M, now 3.1M.
250M operational floor hit
≈ Sep 242026
Top track since the Sep 11 reweight — the corridor-lapse path (50%): draining at 1.35M barrels/day from 286.6M barrels (week ending Aug 28). The 40% standoff path reaches it Oct 19.
Next floor — the 180M operable limit: ≈ Nov 14, 2026 on the lapse path, Jan 27, 2027 on the standoff path.
Crude is the easy half of the story — the scarce thing is the diesel, gasoline, and jet fuel that only refineries make. US refineries are still running above last year's pace; the rest of the world is not; and Russia's refineries are being taken out one strike at a time. The IEA calls the world's refining system “stretched to the limit” in this week's report — Atlantic Basin refining margins hit record levels in August, led by diesel cracks.
Utilization, % of operable capacity, weekly.
The odd one out in this crisis: the US is not cutting, it is pushing at the ceiling. The EIA's weekly report measures runs against operable capacity — in 2026, US refineries have run above 95% every week since Jun 5, peaking at 98.0% the week of Aug 28, and product exports set a record 8 mb/d in August (OPEC). The same stretch of 2025 averaged 90.8%, with a deeper winter maintenance dip. The diesel squeeze is not being made in the United States — it is in Europe, Asia, and Russia, below.
EIA Weekly Petroleum Status Report (WPULEUS3), week ending Friday · utilization = gross inputs ÷ latest reported operable capacity (EIA's definition) · 2025 line = same Jan–Sep window · in mb/d: runs 16.3–17.3 (STEO 4a), above 2025 in every month · the price symptom: the crack-spread chart in Prices.
The squeeze the US hasn't felt yet is the norm everywhere else. Anchors from the IEA's Aug 12 report:
Aug global runs
81.4 mb/d
summer peak, −4.2 mb/d below a year ago (OMR, Sep 11)
FY 2026 runs
−2.6 mb/d
IEA forecast vs 2025 (OMR, Sep 11)
Q3 runs
−370 kb/d
the quarter's further cut (OMR, Aug 12)
estimated capacity remaining, % of pre-strike
The war has a second front: strikes on Russian refineries are destroying the world's capacity to make the fuel other countries import — the diesel and petrol Russia used to sell, and now can no longer spare its own people. The line connects verified anchors only: 100% until the first strikes (Aug 2025), ~75% by mid-April, ~70% by Aug 29 (Moscow Times: over 30% of actual capacity offline). The red bar is the current spread of estimates, because the sources disagree — Ukraine's General Staff says 42.74% is out, Russian Forbes says 54%, the IEA says over 20%.
capacity offline
>30%
Aug 29, Moscow Times — up from ~25% in April
today's estimates: 42.7% out (UA Gen Staff) to 54% (Forbes)
strikes in August
21+
record month, near-daily (Bloomberg, Aug 29)
Kirishi — Russia's #2 plant
halted
~400K b/d, its only NW plant, two strikes in a month (UA.NEWS, Sep 2)
Perm capacity
−86%
primary capacity, satellite imagery (Bloomberg, Aug 25)
every major Lukoil refinery is offline
Novorossiysk — main Black Sea port
hit
fuel-oil terminal + the city, 4 killed (Sep 8–9)
crude outflow 800 → 350 kb/d, Jul → Aug — all three export directions now under attack
stations rationed
28%
nationwide caps; Moscow 90% out of AI-92 (Euronews, Aug 20)
gasoline contracts unmet
>50%
TASS, Sep 3
oil & gas revenue
−45.4%
YoY, Q1 official
the fiscal hit is now as large as the physical one
Russia is banning its fuel exports one at a time — diesel from Sep 30, jet fuel from Nov 30 — hoarding what little it can still make for its own stations, its army, and the coming winter. Each date is the consequence of the damage — a global product-supply shock landing inside the heating season.
Sep 30, 2026
diesel exports
Nov 30, 2026
jet fuel exports
Jan 31, 2027
gasoline & the remaining diesel
And the hole is not being filled: the Belarus rail line — Russia's one external lifeline — is running 25× normal flow, and Kyrgyzstan, which imports over 90% of its petrol from Russia, is down to about six weeks of reserves.
The oil shock is becoming an economy-wide shock — with a lag. The same force that is capping oil prices — demand destruction — is what's slowing the economy. The chart below measures it directly; three channels carry it into the economy: recession risk, a hawkish central bank that is hiking rather than cushioning, and a food shock that lands 12–18 months after the war.
World petroleum & liquid fuels consumption, mb/d · monthly (EIA STEO, Sep 9 release) · Jan – Aug, 2026 vs 2025.
The force that is capping prices, measured directly: world consumption fell up to 4.3 mb/d — about 4% — below last year in May, is still 3.6 below in July, and has recovered to 0.8 below by August. The full year is uglier than that monthly trend suggests: the IEA this week cut its 2026 demand forecast to −2.5 mb/d, from −1.6 in the August edition — the quarterly pace runs −5.3 in Q2, easing to −3.4 in Q3 and −2.0 in Q4, so the year's average is set by a deep first half even as the monthly gap narrows. The destruction is broad: the Middle East itself is the biggest decliner, then the rest of Asia, then the US. The losses are concentrated in middle distillates and petrochemical feedstock, especially in Asia, and the IEA sees demand recovering by 2.6 mb/d in 2027. The institutions disagree on the scale: OPEC, reporting a day earlier, sees 2026 demand up 0.4 mb/d — a 2.9 mb/d spread between the two forecasts.
EIA STEO Table 3e (Sep 9 2026, forecast completed Sep 3) · Jan–Aug 2026 are actuals in that release · world/regional values are EIA estimates (apparent consumption, incl. refinery fuel & bunkering) · Sep 2026 onward is forecast, not shown.
Million b/d, July 2026 vs July 2025 — same table.
World
−3.6
105.4 → 101.8 mb/d
Middle East
−1.2
10.2 → 8.9 mb/d
Asia & Oceania
−1.9
37.9 → 36.0 mb/d
China
−0.8
16.4 → 15.6 mb/d · of Asia & Oceania — holding up on stockpiles
Europe
−0.1
14.8 → 14.7 mb/d
United States
−0.4
21.0 → 20.5 mb/d
Odds of a US recession (July–Sep 2026). The banks' numbers are for the next 12 months: 25% to 50%, with Goldman up from 20% to 30% since the war began. Polymarket's number is for a longer window — recession by end-2027, about 15 months — and it is 32%. That matches the banks' number despite the extra three months. The consensus still expects to dodge it; the banks and the betting markets say otherwise.
The standard recession-aversion move — cutting rates — is off the table. A newly hawkish Fed is pricing in hikes. The Aug PPI printed 5.4% on Sep 10; the Aug CPI came out Sep 11 — headline 3.4% (flat), core 2.4% (easing), energy +16.3% for the year. A hike into the winter drawdown would be the single most bearish macro signal in this model.
Fed funds, July FOMC
3.50–3.75%
9-to-3 hold; officials 'see the need for a hike if inflation doesn't cool'
Odds of a HIKE in September
≈86%
fed funds futures, Sep 11 — up from ~72% Thursday after the Aug CPI core ran above forecast; Polymarket 62%
Aug PPI (BLS, Sep 10)
5.4% YoY
+0.4% m/m, in line; up from 4.8% (Jul, rev.) — energy +4.2% (diesel +24.1%); 10-yr at highest since Oct 2023
Meanwhile the ECB says Germany and Italy would both be in a technical recession by end-2026 if the conflict drags on — the European version of the same trade, without the Fed's balance-sheet cushion.
The oil shock reaches the rest of the economy through two numbers: what the government pays to borrow, and what you pay at the register. Both are at multi-year highs, and together they explain why the Fed is talking about hiking rates instead of cutting them to cushion a slowdown.
10-yr US Treasury, % · weekly closes (FRED) · Jan 2 – Sep 10.
The 10-year is the rate the US government pays to borrow for a decade — the benchmark behind mortgage and business loan rates. It is up 100 basis points from before the war. Two forces are pushing it up: inflation stuck above 3%, which keeps the Fed hawkish (a 25bp hike on Sep 16 is now priced at about 86%, against 62% on Polymarket), and fiscal worry — national debt above $40 trillion with $8.4 trillion of Treasuries rolling over by year-end. The 4.8% line — the high reached in January 2025 — has now been broken and held: 4.83 on Sep 9, 4.95 on Sep 10, 4.97 on Sep 11 with an intraday touch of 4.992 — the highest since October 2023, the level strategists say would start causing "meaningful problems" in other markets if it holds (CNBC, Sep 11). The 2-year sits at 4.63%, its highest since July 2024 — that's where the hike is priced. The 10-year breakeven (inflation expectations) actually eased to 2.36%, so the sell-off is a real-rate story, not an inflation-expectations story.
CPI (retail, amber) + PPI final demand (wholesale, blue) · % year-over-year · monthly (BLS) · Jan – Aug.
The consumer side of the crisis: inflation jumped from about 2.4% to 4.2% in three months as the oil shock hit the pump, then eased to 3.4% in July — and held exactly there in August — +0.4% for the month, with energy up 2.1% (gasoline +3.9%, over a third of the increase) doing the work, and core easing to 2.4% from 2.5% — though the monthly core rise (+0.3%) still ran above forecast, and it was the last major print before the Fed's Sep 16 meeting. Still well above the Fed's 2% target. Wholesale prices ran hotter and earlier: the PPI peaked at 5.9% in May and printed 5.4% in August (July revised to 4.8%) — energy, with diesel up 24.1% from a year ago, explains most of the rise. Factories feel the oil shock before households do (BLS, Sep 10–11).
Gas is about a third of the marginal cost of nitrogen fertiliser. With Qatari LNG offline for 3–5 years and Asian gas at all-time highs, the 2027 harvest is a one-shot bet: farmers plant in March–April on the fertiliser prices set that winter. The last two times fertiliser plus freight spiked like this — 2007–08 and 2022 — food prices followed within 12–18 months. There is no model behind this one: unlike the charts above, these are documented lag windows from two historical shocks — order-of-magnitude timing, not a forecast.
Gulf → India VLCC freight
+411%
$4.34/bbl in Aug vs pre-war (Frontline)
TTF gas (Europe)
≈€50/MWh
Aug 19 · ≈$16.8/MMBtu
JKM gas (Asia)
$23.17/MBtu
all-time high
The lag chain: energy shock → freight + gas → fertiliser → plantings → harvest → food prices. Each bar is the documented window from 2007–08 / 2022. The spike lands on top of the terminal state, not before it.
Dated events that will tell us where this is heading.
Any day
Saudi announcement on the East–West pipeline — the "precautionary suspension" after Thursday's attacks (now attributed to drones from Iraq's Maysan province): a restart date, or a damage assessment
the bypass is what keeps the standoff track alive — after today's reweight its status is the single most-watched variable
Sep 14
Gulf foreign ministers meet Iran in Oman — the war's first Gulf–Iran ministerial; the Iran–Oman shared-route agreement is to be signed and notified to the IMO. Same day: the US Treasury designates a major Iranian bank
the holds branch's reconstitution path finally has a date — though FT-cited diplomats say a full reopening still needs a US–Iran understanding, which does not exist
Sep 16
The Fed's decision — a 25bp hike (3.50–3.75% → 3.75–4.00%) is priced at ~86% (CME; Polymarket 62%)
August CPI's core came in above forecast, the last major print before the meeting; the 2-year is at 4.63%, its highest since Jul 2024, and the 10-year touched 4.99% on Sep 11 — highest since Oct 2023
Sep 16
EIA weekly report (week ending Sep 4) — do US diesel stocks fall under 100M barrels, and how fast is the SPR being drained now?
the earliest sign of whether things are worsening or easing
Sep 30
Four things in one week: Russia's diesel export ban expires (already extended to this date — the choice is extend or open), the US-led coalition withdrawal from Iraq is due to complete (Patriots leaving Erbil), the Sep 30 prediction-market bets settle, and Sweden's fuel-tax cut ends
the densest cluster of dated events
Oct 7
The EIA's monthly outlook — the first one written after the tanker war, the refinery strikes, and the no-sailing zone
the real test of the EIA's "constrained, but open" view — and of the EIA–market spread (EIA 2H26 ~$90 vs settled spot $104.61 — a ~$15 gap)
Nov 30
Russia's jet-fuel export ban takes effect
the start of the airline fuel crunch
The order in which the world runs out, if the crisis continues. The first two dates are arithmetic on published stock levels; the rest are inferences from customs and stock data — expect them to be rough by a week or two.
US East Coast
Sep 14–21
diesel and heating-oil stocks fall below a month of supply — already 27% lower than last year
Russia
Sep 30
the diesel export ban runs out — refineries are 30%+ damaged, so Russia simply has no diesel left to sell; survival, not politics
Europe's oil hubs
Early Oct
Rotterdam-area diesel stocks drop below the level where traders start panicking (8.5–9M barrels); mid-October if the strait closes fully
China
≈ mid-October
commercial buffer stocks start draining faster than normal
Europe, at the pump
≈ late October
shortages reach ordinary consumers — and governments buckle under price spikes first
Air travel
Nov 10–30
Russia bans jet-fuel exports (Nov 30) — the world's remaining buffer is ~26 days of flying