Day 103. The US-Iran exchange that began with the Apache loss over Hormuz is probably closed, not confirmed. CENTCOM’s ~20-target package was the largest single-night strike of the crisis (prior responses ran 2-4 targets) but calibrated: sensors and command nodes of the IRGC Hormuz kill chain, no leadership, oil, or nuclear targets, all inside a four-hour window capped by “operations complete,” a deliberate cycle-termination signal. Iran’s answer, drones at Bahrain and Kuwait and ballistic missiles at Jordan with zero confirmed casualties anywhere, matches its June 6 honor-satisfying close-out. Across the June 3/6/10 cycles: 25+ ballistic missiles, dozens of drones, zero confirmed casualties or facility hits; Iran’s claim of 21 facilities hit is denied by hosts, assessed propaganda. (Trump called the Apache “shot down”; a midair collision has not been ruled out.)

One variable decides whether the cycle stays closed: does Washington absorb the June 10 counterstrike. The IRGC’s June 6 tripwire is on record: further US strikes on Iranian territory mean moving to “completely close” Hormuz, so a second US strike night produces a closure decision point, not another symmetric cycle. Full war resumption moves to 18-24% from 12-13%, front-loaded into the next 72 hours, decaying to 14-16% if absorption holds.

Jordan is the politically loaded move. Muwaffaq Salti sits ~1,200 km west of every prior Iranian aim point, and Iran sent ballistic missiles at the newest theater while drones went to the familiar ones: deliberate pressure on Amman, the weakest political link among US hosts. Iran’s foreign minister made it doctrine: regional nations bear “legal and moral responsibility” to prevent US strikes from their soil. Watch for caveats, not ruptures: 40-50% odds at least one host imposes a basing or overflight caveat within two weeks; rupture under 5%. Caveats push US options toward carrier aviation and standoff weapons, which are harder to calibrate, and the bounded-cycle grammar degrades.

A second structural worry: the counterstrike’s profile (execution within hours, pre-surveyed aim points, one pre-approved escalation branch at Jordan, uniform weaponeering) fits pre-delegated IRGC launch authority. Assessed, not confirmed; tied to the unverified report of Khamenei’s comms cut since June 8. Pakistan’s two letters for Khamenei sit unacknowledged three days on. If real, there may be no counterparty able to sign a deal, and a signed one might not stop the protocols. Discount Trump’s “two or three days” deal prediction: fifth timeline signal in four weeks, prior four missed.

Brent closed June 9 at $91.11 (-3.42%), then flat near $91 through June 10; WTI $87.95. The shrug is rational on the modal path: Hormuz already ran at 2% throughput, no new flow delta, legible off-ramps. What $91 misprices is the distribution: war odds nearly doubled overnight. Beneath the premium, supply tightens: API -9.1M bbl (seventh weekly draw, ~420M bbl operational floor 1-2 weeks out once EIA confirms) and China importing ~7.8 mbpd, an eight-year low. Probability-weighted fair value on current odds is $105-112; spot is cheap to the tail. Levels: $90 support, $93 repricing trigger, $97.68 breakout confirmation, $85 hard floor. Markets like this gap; they do not trend.

Hormuz risk changed shape, not size. IMF PortWatch counted 2 transits on June 7 against a 94/day baseline, down from the stale ~10/day figure; Energy Secretary Wright says exports are rising, and we carry both. Degraded IRGC sensors cut precision-interdiction risk but raise the indiscriminate kind: mines, visually targeted fast-attack swarms, misidentification. Marivex (June 8) fixed enforcement doctrine: warnings, engine-room disabling shot, third-party crew rescue. Seventh vessel since April 13, 122 redirected as of early June; false registries and unladen status protect nothing; all 24 crew were Indian. VLCC war risk runs ~8.0x pre-crisis (straits.live); Lloyd’s List puts the war-risk cost at $10-14M per Hormuz voyage.

Red Sea exposure is the deceptive one. MARAD 2026-006, ~48 hours overdue, still claims no Houthi attacks since October 2025; the water is an active anti-ship ballistic missile zone with three strikes in 48 hours and no US carrier within reach (both carriers sit on the Hormuz axis; none has crossed Bab el-Mandeb since December 2023). Houthi targeting keys on corporate beneficial ownership, not flag. Cape routing at $130-145K/day is now the default, adding 12-15 days and $1.6-2.2M in hire. JWC listed-area expansion is days away. Suez/SUMED (~5.5M b/d) falls next as June fixtures divert. One Houthi strike on a vessel with zero Israeli affiliation converts enforcement into blanket closure.

Calendar pressure is fixed. The IEA window closes July 1, OPEC+ meets July 5, Saudi SPR exhaustion tracks to ~July 19; only ~half of the +188K bpd July increment can physically reach water, and Saudi’s share moves only via Yanbu under live Houthi threat. Iran-Israel held overnight (IDF struck Tyre, 5-8 killed per the Lebanese Health Ministry, but not Dahiyeh). Watch: any further US strike on Iranian territory; Amman and Kuwait City basing statements; any Iranian strike producing casualties; the MARAD update; a Brent close above $93.