Brent fell 19% in May — the worst month since COVID — while Hormuz ran at 5% of normal, Iran's exports hit a six-year low, and the SPR drained 60 million barrels. The entire collapse rests on a 60-day memorandum that Trump has not signed. At 5:00 PM tonight, after the close, CENTCOM struck Iranian air defenses on the strait. Brent settled at $91.45 before the missiles flew.
The SPR release ran ~1.27 mb/d in May, not the modeled 1.43 (the IEA counts only ~31 Mb sold under the US program by May 8 — definitions differ; both are right). The "approaching $800K/day" VLCC figure is softened: best-documented peaks are $423.7K (Mar 2 record) and ~$474K (Apr 17). War-risk premium sources conflict by an order of magnitude (S&P: ~1% of hull at peak, 0.3–0.5% post-ceasefire; Lloyd's List: $10–14M/voyage); we now print direction, not level. The French ship struck in the May escalation was hit May 5, not May 1. And №03's "SPR ~409M" anchor was the April 10 reading — by its April 28 publication date the SPR was already ~398M (the 2026 peak was 415.4M in early March).
Act I (May 1–15): Iran turns Hormuz into a tollbooth and prices grind to $109. Act II (May 18–29): a 60-day peace framework emerges and Brent loses 19% in a month — the worst since COVID. Act III (June): the deal sits unsigned, the blockade never pauses, and tonight the shooting resumed.
The shape matters as much as the level. The grind to $109.26 happened on tanker seizures and a tolls-and-uranium deadlock; the collapse to $92 happened on headlines about a deal that does not yet exist. Two of the window's biggest down-days (−5.3% each) came on "optimism." Physical flows barely changed the whole time.
On May 5 Tehran stood up the Persian Gulf Strait Authority — apply by email, disclose ownership and cargo, pay $1–2 million per transit, settled in yuan. It went live May 18. India, Iraq, and Pakistan negotiated carve-outs. The US, EU, and Gulf states call it illegal. It is also, functionally, the only way through.
Beneath the toll regime runs a quieter channel: a JPMorgan note dated June 4 estimates ~2 mb/d is leaving the Gulf on tankers with transponders switched off — with the US Navy quietly coordinating some exits. That dark flow, not the PGSA queue, is why Energy Secretary Chris Wright could tell CNBC on June 9 that Hormuz traffic is "rising very meaningfully" — the remark that catalyzed the day's 3% slide. Meanwhile the France–UK multinational escort mission is still not operational (the Charles de Gaulle group and HMS Dragon are in theater, Italy has committed a four-ship task force with minesweepers — all conditioned on a stable ceasefire), and the diplomacy is now hostage to a third front: the MOU stipulates an end to the Israel–Hezbollah war, and on June 3 Iran threatened to stop communicating and "completely block" the strait over Israeli operations in Lebanon. Bloomberg's three sticking points — Hormuz, Lebanon, enrichment — are all still open.
Vol. I №03 published a full model on April 28. Five weeks of data are in. The base-case Brent call was nearly perfect — for the wrong reasons. Flows tracked our pessimistic case. Price tracked our base. That gap is the most important lesson in this dispatch.
| №03 prediction (Apr 28) | Actual | Verdict |
|---|---|---|
| Brent May avg: base $108 | $107 (EIA) | ✓ Nailed |
| Hormuz May: 22 vessels/day, 5 mb/d (base) | ~6/day, ~3 mb/d | ✗ Below even pessimistic |
| Gasoline peak $4.27, week of May 25 | $4.56 on May 21 | ✗ 29¢ low, timing right |
| California $6.35–6.45 in June | Peaked $6.16 May 7 · $5.84 now | ✗ Overshot |
| SPR May 31: 387M | 357M (May 29) | ✗ Drained 30M faster |
| Fed: hold, first cut September | Warsh chairs · ~70% Dec hike priced | ✗✗ Regime flipped |
| JPM May 9–30 window: no national crunch (base) | Peak $109.26 · no spike · Cushing −6 wks to 22.4M | ✓ Right; regional stress real |
| April CPI 3.7% | 3.8% | ✓ |
| No recession in Q2 (65%) | GDPNow +3.3% · payrolls +172K | ✓ |
| Shale response "+0–50 kbpd by Q1'27" | 13.8 mb/d output · rigs +19 in 6 weeks | ✗ Shale is answering $95 |
The April model mapped flows to price too directly. Flows came in at the pessimistic branch; price came in at base — because four buffers we under-weighted absorbed the difference. Those buffers are now half-spent.
— Lesson of the May windowJPMorgan's "operational minimum" window came and went at $109, not $150. Its own June 5 verdict: inventories drew slower than projected and demand destruction proved larger than expected. Four buffers — roughly 10+ mb/d combined — stood between pessimistic flows and pessimistic prices.
These buffers muted the upside once. Demand is already destroyed; the SPR is half-spent (EIA's own STEO projects 243.5M bbl by Q3); commercial stocks sit 3% below the 5-year average with Cushing 2.4M above its floor. If the pessimistic branch fires now, the same flow shock meets far thinner shock absorbers. That is the central risk architecture of the next 90 days.
While Brent fell 19%, US commercial crude fell six consecutive weeks (−30.7M bbl), the SPR posted its 11th straight draw, refineries ran at 95.3%, and Cushing closed within 2.4M bbl of its operational floor — the squeeze that briefly put WTI above Brent for the first time in the war.
| Week ending | Crude (Mb) | Δ w/w | Gasoline | Distillate | Cushing | SPR | Refinery % |
|---|---|---|---|---|---|---|---|
| May 1 | 457.2 | −2.3 | 219.8 | 102.3 | 29.1 | 392.7 | 90.1 |
| May 8 | 452.9 | −4.3 | 215.7 | 102.5 | 27.4 | 384.1 | 91.7 |
| May 15 | 445.0 | −7.9 | 214.2 | 102.9 | 25.8 | 374.2 | 91.6 |
| May 22 | 441.7 | −3.3 | 211.6 | 100.8 | 23.0 | 365.1 | 94.5 |
| May 29 | 433.7 | −8.0 | 215.0 | 102.3 | 22.4 | 357.1 | 94.7 |
| Jun 5 | 426.5 | −7.2 | 215.1 | 102.1 | n/a | 349.2 | 95.3 |
All values from EIA WPSR primary releases. Distillate ~13% below 5-yr average (EIA's Jun 3 "3%" is an apparent typo — flagged). Distillate never breached 100M — the low was 100.8M on May 22. Total products supplied, 4-wk avg: 20.4 mb/d, +3.0% y/y — aggregate US demand has not cracked.
№03's scenarios keyed on the strait. The market has moved on: the binding variable is whether Trump signs. Rebuilt scenarios — Signed & holds (25%) · Stalemate grind (45%, base) · Re-escalation (30%) — with flows as the slow variable and buffer depletion as the asymmetry.
| Month | Signed (25%) | Base (45%) | Re-escalation (30%) | Prob-Wtd |
|---|---|---|---|---|
| Jun '26 | $93 | $95 | $99 | $96 |
| Jul | $88 | $94 | $115 | $99 |
| Aug | $84 | $92 | $122 | $99 |
| Sep | $80 | $90 | $118 | $96 |
| Oct | $77 | $88 | $112 | $92 |
| Nov | $75 | $87 | $108 | $90 |
| Dec | $73 | $86 | $105 | $88 |
| Jan '27 | $72 | $84 | $102 | $86 |
| Feb | $71 | $82 | $100 | $85 |
| Mar | $70 | $80 | $96 | $82 |
Scenarios weighted 25/45/30. Heating oil materially below №03's path — the crude collapse passed through before the heating season. Rockets-and-feathers retained: declines pass at ~55% over 4 weeks. CA gasoline carries a July 1 LCFS wildcard of +$0.15–0.65/gal.
The April model's most important carry-forward: after Q3 the SPR is no longer a policy tool. A re-escalation in Q4 meets a buffer roughly half the size it was in March — with OECD commercial stocks already 400+ Mb lighter and a restocking bid waiting underneath any price decline.
| Month | Vessels/day | Oil (mb/d) | Constraint |
|---|---|---|---|
| Jun '26 | ~6 | 3.0 | Toll regime + carve-outs only |
| Jul | 8 | 3.5 | Stalemate; episodic strikes |
| Aug | 10 | 4.2 | Carve-out creep (India · Iraq · Pakistan) |
| Sep | 14 | 5.5 | Partial demining if MOU progresses |
| Oct | 18 | 6.5 | Insurance still multiples of pre-war |
| Dec | 25 | 8.5 | Bypass (9–10 mb/d) does the heavy lifting |
| Mar '27 | 35 | 10.5 | Full normal requires the signed deal |
The structural shift from №03: the bypass network, not the strait, is now the main supply variable. Basra–Haditha broke ground May 1 (2.5 mb/d design); the UAE's second Fujairah line is ~50% complete and accelerated to 2027; Kirkuk–Ceyhan targets 600 kb/d. Every month of stalemate de-Hormuzes the map.
National gasoline peaked at $4.56 on May 21 — a four-year Memorial Day high — and has since shed 40 cents in three weeks, the fastest decline of the war. California peaked at $6.16 in early May and is back under $6. The rockets flew in April; the feathers are falling now.
| Date | National regular | California | National diesel (EIA) |
|---|---|---|---|
| Apr 30 | $4.30 | $6.01 — first >$6 since 2023 | ~$5.55 |
| May 7 | $4.55 | $6.16 — the peak | — |
| May 21 | $4.56 — the peak | $6.14 | — |
| May 28 | $4.42 | ~$6.05 | $5.52 |
| Jun 4 | $4.24 | $5.97 | $5.35 |
| Jun 9 | $4.16 | $5.84 | $5.21 |
| Sep (base fcst) | $3.70 | $5.52 | $4.82 |
| Dec (base fcst) | $3.50 | $5.32 | $4.70 |
| Mar '27 (base fcst) | $3.40 | $5.22 | $4.55 |
No rationing materialized — the California Energy Commission saw "no supply shortfall within six weeks," and spot outages stayed local. Newsom again rejected a gas-tax holiday. The state's July 1 LCFS update (+$0.15–0.65/gal) is now the single biggest variable in West Coast prices. Benicia ceased refining on schedule; Valero is importing into NorCal to fill the ~75 kb/d gap.
№03 modeled a September cut. Instead: April CPI hit 3.8%, PPI hit +6.0% y/y (most since 2022), May payrolls beat at +172K — and the market now prices ~70% odds of a December hike. Kevin Warsh chairs his first FOMC June 16–17. Tomorrow morning's May CPI (consensus 4.2%) lands twelve hours after this dispatch.
| Month | Headline CPI (base) | Core | Food-at-home | Driver |
|---|---|---|---|---|
| May '26 | 4.2% (peak) | 2.9% | 3.0% | April gasoline in the index |
| Jul | 3.7% | 3.1% | 3.2% | Freight + fares pass-through |
| Sep | 3.3% | 3.0% | 3.4% | Energy fading, food arriving |
| Dec | 3.2% | 2.9% | 3.9% | Food channel peak window |
| Mar '27 | 2.9% | 2.8% | 3.8% | Base effects roll off |
The transition itself was contentious: Warsh was confirmed 54–45 on May 13 (Fetterman the lone Democratic yes) and sworn in May 22; Powell stays on the Board as a governor; the April FOMC drew four dissents — the most since 1992, including Gov. Miran wanting a cut even as markets price hikes. Base case: the May crude collapse delivers energy disinflation that prevents the priced December hike. Re-escalation case: headline re-peaks at 5.3–5.8% in Q4 and the first hike arrives — into a slowing economy. That is the 1979 sequence.
The UAE's OPEC exit took effect May 1 — the first Gulf producer out. Kuwait remains in force majeure; Ras Laffan's force majeure was extended through mid-June; Iraq's south still exports near zero by sea while Ceyhan and trucking carry ~230 kb/d. Iran's economy is the war's quiet decider: GDP −6.1%, inflation 69%, food inflation 105%, the rial at 1.32M, fuel rationed at 10–15 liters. Mojtaba Khamenei has not appeared in public since March. Tehran is negotiating because it is losing — which is precisely what makes its strait leverage existential, and the sequencing stalemate so durable.
№03 asked which market was wrong — equities at records or sentiment at record lows. May's answer: both kept going. The S&P melted up to ~7,600 through the worst oil month since COVID while UMich made a new all-time low and the Senate (50–47, May 19) and House (215–208, June 3) passed the first war-powers checks of the war. The political economy of $4.16 gasoline in a midterm year is now doing what the bond market hasn't: constraining the war.
| House | The call | What happened | Verdict |
|---|---|---|---|
| JPMorgan | May 9–30 "op-min" → non-linear spikes, path to $150 | Peak $109.26; admitted demand destruction beat its model (Jun 5); now $96 avg '26 | Spike wrong |
| Goldman | Q4 $90 (4th upgrade) | Holding; +1 month closure → $120 Q3 scenario | In the money |
| Citi | $120 near-term · $150 bull | $109 peak, then −19% | Too hot |
| Capital Economics | $130–140 "next month" (May 16) | June MTD avg ~$94.7 | Falsified |
| EIA STEO | $105 Jun–Jul · $95 avg 2026 · $89 Q4 | Assumes closure, no peace discount — June running $10 below | The flow-priced view |
| OPEC MOMR | 2026 demand still growing +1.2–1.4 mb/d | IEA −0.4 / EIA −1.1 — a 2.5+ mb/d agency gap, the cycle's widest | The outlier |
The market is trading $10–15 below the flow-consistent (STEO) path — that gap is the peace discount, and it is the single number to watch. A signature closes it from below; a broken ceasefire closes it from above.
May proved the market will price diplomatic probability over physical scarcity — flows at the pessimistic branch, prices at base. The corollary: when the diplomacy breaks, the repricing is violent and instant. Tonight's strikes are the first test.
SPR 409→349 and headed for ~250–310 by fall. OECD stocks −400+ Mb. Demand already destroyed. The same shock that produced $109 in May produces materially more in October — with a restocking bid under every dip.
A state now charges yuan-denominated tolls on an international strait. Basra–Haditha broke ground; the UAE's second bypass is half-built; 7.8 mb/d already flows around. Whoever wins the war, the strait's monopoly is ending — a structural cap on the next crisis's premium.
Every dollar of the May collapse was priced on a signature that does not exist. At five o'clock tonight, the United States struck Iran again. One of those two facts will give way first.
— Bottom line · June 9, 2026, 23:00 ET