Oil Crisis — Living Summary ¶
Last updated: Sep 9, 2026 (Day 193) — big pass: the war's biggest tit-for-tat wave (10 Iranian tankers gone in a week; Iranian missiles on a US base in Jordan), Iran's exclusion zone enforced before declared, Brent back above $100 for the first time since Jul 24 (user report verified), and the Sep 9 EIA STEO landed. Branch weights reweighted (regime event). ~2-minute read. The main report has the full analysis; the Model Card is the compact spec. Plain-language version for non-readers: Oil Crisis — Plain Summary.
Where things stand ¶
Two chokepoints closed (Hormuz + Red Sea/Bab el-Mandeb), a US–Iran tanker war now running on a published exchange rate (3 Iranian tankers per 2 warships targeted — 10 Iranian tankers destroyed/disabled in one week, Sep 5–9), Iran's "exclusion zone" enforced before formally declared (most strait traffic now rides the Iranian-controlled route; Kpler ~10 transits/day, down from 13–15), and the war's first strike on a third country (20 missiles on the Al-Azraq US base in Jordan). The refinery front widened: Houthi strikes re-hit Aramco's Jazan refinery (offline since at least late July, nameplate ~400 kb/d), and Ukrainian drones hit the Novorossiysk export terminal — the Russian strike campaign now reaches the export side. OPEC+ has paused four consecutive production hikes (Oct flat at 31.01M bpd) — spare capacity is trapped, not withheld.
The scarce asset is product (diesel, jet, East Coast diesel), not crude — and the EIA's own STEO now says so (distillate crack +11.6%; US distillates forecast below the 5-yr low "through much of 2027"). US retail diesel set a fresh all-time record ($5.94, Sep 9); gasoline $4.22 (+10¢ in a week) — the political-break tripwire keeps moving further past.
Key numbers (prices Sep 9; EIA WPSR w/e Aug 28) ¶
| US SPR | 286.6M bbl (lowest since Dec 1982), drawing ~0.45M b/d — inside the ~300M cavern floor; ~82 days to the 250M operational floor at this pace. Next print Sep 10. |
| US distillate | 104.2M bbl, −14% vs 5-yr avg; EIA STEO: below 100M in Sep, below the 5-yr low through much of 2027; PADD1 the weak spot, −27% YoY |
| US gasoline | 205.7M — safe in all branches (net exporter) |
| ARA gasoil | 12.07M bbl (4-yr low); trader floor ~8.5–9M crossed ~late Oct |
| Russia | >30% of refining offline; Kirishi fully halted; all three export directions now under attack (Novorossiysk fuel terminal hit — 4 killed — plus CPC and Ust-Luga/Baltic); Black Sea crude outflow 800→350 kb/d Jul→Aug; "second wave" domestic crisis: 17 regions rationing, gasoline at ~70% of consumption, importing from India; producer-diesel ban extended to Sep 30 (T-21) |
| Asia | Japan 203d, Korea >1y (parliamentary consent motion for Hormuz ops could come this month), China buffer untouched (est. 1.2–1.4B bbl); China's commercial buffer runway expires Q1–Q2 2027 |
| Prices | Brent $100.71 (Sep 9; >$101 intraday — first since Jul 24); WTI $96.38; AAA regular $4.2245; diesel $5.94 ATH (CA $7.87 ATH); Goldman: $120 "plausible" |
| EIA STEO (Sep 9) | Brent 2H26 ~$90, 2027 $74 (from $69); Middle East below pre-conflict output until 2Q27; global stocks −400M bbl YTD. Inputs froze Sep 3 — does not price the tanker war; ~Oct 7 STEO is the next real signal |
| Traffic | Kpler 10 transits/day (most on the Iranian route); Lloyd's 102/week vs ≥130/day prewar; Red Sea dry cargo partly back, crude tankers ❤️/day |
The base case ¶
Standoff drift (~50% — reweighted down from 50–55%): the standoff is now a hotter one — both sides' stated rules are being executed rather than announced (a published US exchange rate, applied twice in four days; an Iranian exclusion zone enforced "by claim and by missile before any document exists"). Corridor holds: ~10% (the lanes now run inside the exclusion zone, mostly Iranian-controlled); corridor lapses: ~40% (up from 30% — the regime event of Sep 8–9). The Oman safe-corridor talks (IMO filing claimed, unverified) are the swing factor: if real, the corridor reconstitutes under Iranian management.
The headline is unchanged: the status quo cannot last 12 months. By month 9, either the corridor or China's buffer breaks — and that break (re-import into a broken supply base, or a 1.2B-bbl release) is the largest market event of the war. The Sep 9 escalation compresses the dates another ~1–3 weeks; the sequence is untouched.
What decides it, in order: (1) Russia's Sep 30 ban (T-21 — extend into a 17-region shortage or lift it; landing on a base that now includes Novorossiysk, CPC and Ust-Luga); (2) the corridor window — now dominated by whether the Oman corridor gets IMO-registered or the exclusion zone hardens; (3) ARA trader floor + first EU price caps (late Oct–Nov); (4) SPR at 250M (late Oct–Nov, pace-dependent); (5) Russia jet ban (Nov 30). Plus the one undated risk: a calibrated Russia→Europe hybrid attack that fails to fail.
Even the good case is an 18-month recovery (ADNOC: full Middle East flow unlikely before late 2027; EIA now says Middle East below pre-conflict output until 2Q27). And the baseline itself is damaged — reopening does not restore pre-war supply (Qatar LNG −17% until 2029–31; crude −0.6M bpd structural, compounding to 1.5–2.5M with drift).
Watch list (next two weeks) ¶
- WPSR Sep 10 (w/e Sep 5) — first inventory print after the escalation; the distillate sub-100M breach and the SPR draw pace are the lines to check. Aug PPI the same day — consensus 5.4% from 4.7% is the macro tripwire.
- The Oman corridor IMO filing — filed or not? It decides whether "corridor holds" reconstitutes under Iranian management.
- Exclusion zone coordinates — "to be announced"; when published, check them against the shadow southern lanes.
- Iran's tanker threats at Kuwaiti/Bahraini berths — stated, untested; a strike at berth would be a new regime event and reweight immediately.
- Jazan restart — the repair clock started before Sep 8 (restart already slipped to late August); every week offline is a point toward permanent partial loss.
- Polymarket Sep 14/15 settlement ("Hormuz normal by Sep 30", 3.8%) — expected ~0.
- US political response to the $5.94 diesel record + PPI — the question is how many weeks after the record.
Updated at the end of every research pass. Index: Oil Crisis — Research Index.
.svg){kind=logo}