Geopolitical Deep Dive
Updated July 20, 2026

Voyager Research · Geopolitical Deep Dive · Strait of Hormuz

Hours to Break, Months to Mend

Iran cannot seal the Strait of Hormuz, and has not needed to. Functional closure, one cargo class and one time-window at a time, is the durable weapon.

Independent analytical research. Not personalized investment advice.

Assessment. Four days ran the experiment. The de-mining window the June MOU promised lapsed on July 17 with no clearance begun by any account located; Iran’s deputy foreign minister said on July 18 that Tehran had suspended its MOU commitments, and Washington answered with a shrug. Re-imposition had taken hours the week before; nothing about reopening advanced since. Brent settled 5.8% higher on the week, and a thin weekend spike toward $91 was flat by Monday morning, so the market still prices disruption as a chronic tax. Six of seven July 16 Soundings fired, five with the thesis. The revise conditions stand unmet: certified clearance and a surviving settlement remain absent, and Iranian export economics have not broken.

The Fix. The position, one line per judgment:

1. The thesis survived contact: Iran re-broke the strait in hours in the week to July 14, and the four days since produced no step toward reopening.
2. Diplomacy is idle in practice: the de-mining window lapsed unmet around July 17 and Iran says it has suspended its MOU commitments (July 18); neither capital still mentions August 16.
3. Six of the seven Soundings pre-registered on July 16 fired within four days; The Log below grades them, including one that fired against the monetization read.
4. The market repriced without breaking: ICE Brent settled at $88.10 on July 17, up 5.8% from the July 13 settle, and a thin weekend run to $91.41 had faded fully by Monday morning.
5. Iran’s chokepoint revenue stalled: its transit-fee authority suspended passage on July 12, and the last US oil waiver expired on July 17.
6. The blockade is biting: CENTCOM’s ledger reached six vessels redirected and one disabled as of July 19, the first sustained enforcement against Iran-bound traffic since April.
7. The data is degrading faster than the strait: PortWatch’s public series stops at July 12, and trackers confirm the zero-LNG streak only through July 17; later readings rest on labeled secondary sources and belligerent claims.

Positioning: who gains, who loses

Gulf producers with bypass pipelines gained: Saudi Arabia ran its East-West line to a record and is diverting roughly 4 million barrels a day through Yanbu (July 2026, trade press). Iran, at least through the truce months, gained more than the priors suggested: its tension-month oil revenue ran above its early-February run-rate on both major tracker bases, and it has charged for passage since the war’s first weeks. That gain paused on July 12, when the toll authority suspended its own passage regime; the last US oil waiver died five days later. Tanker owners and war-risk underwriters captured the premium on every voyage that did move. Qatar lost most among producers: LNG has no pipeline bypass, and the Ras Laffan strikes sidelined roughly 17% of its export capacity for years. Kuwait joined the losers on July 18: its state oil company reported significant material damage and injuries at an oil facility from repeated Iranian attacks, and its authorities reported strikes on two desalination plants. The buyers of last resort lost quietly: Asia takes 89% of Hormuz crude, and Japan’s imports fell by roughly half. India removed its seafarers from the trade by regulatory order on July 16. Seafarer deaths and injuries from the tanker strikes are recorded in the IMO’s incident listings.

The Log

The July 16 edition pre-registered seven Soundings. Four days later, the grades:

The strait as a bill of flows

Start with what the waterway carried before the war, because every percentage in this piece hangs on it. The EIA’s chokepoint series puts first-half 2025 flow at 20.9 million barrels a day of oil (14.7 crude and condensate, 6.1 refined products) plus 11.4 billion cubic feet a day of LNG, about one fifth of global LNG trade, primarily Qatari. The IEA’s larger figure of roughly 24 million barrels a day measures total Gulf exports including the bypass pipelines, a distinction that matters later. Vessel counts come on a ladder of scopes: the IMF’s PortWatch satellite series averages about 90 commercial transits a day pre-war (tankers half of that), Kpler counts ~100, and Lloyd’s List Intelligence counts ~130 on its own scope (cargo carriers of 10,000 deadweight tonnes and up). This piece states vessel percentages on the PortWatch spine and volume percentages on the EIA spine, never mixed.

The demand side is concentrated where the risk tolerance is highest. China took 37.7% of Hormuz crude and condensate in early 2025; China, India, South Korea, and Japan together took about three quarters. Japan runs roughly 90% of its crude imports through the strait, and its imports fell by nearly half in the March-May war months.

Five months of war, in transits

The campaign opened on February 28, 2026: Operation Epic Fury (the US designation, per CENTCOM’s own operation page) alongside Israel’s Operation Roaring Lion. Two days later the first formal Iranian closure declaration came from IRGC senior adviser Ebrahim Jabari: the strait was closed, and violating ships would be “set ablaze.” Transits collapsed within seventy-two hours: by late March 2 the tracker Windward recorded zero active tanker transits, and on March 3 only four vessels of any kind crossed, against a seven-day average near 77 before the war.

What followed settled into a cycle. A ceasefire arrived around April 7-8; talks in Islamabad collapsed on April 12; the US imposed a naval blockade of Iranian ports on April 13, and Iran answered with re-closure; Washington announced escorts on May 4 (Project Freedom, paused within 48 hours after two escorted transits, then quietly resumed by June per Bloomberg reporting); a memorandum of understanding was signed on June 17 with a 60-day negotiating window; strikes resumed within weeks, and the truce died on July 8-9. Iran declared the strait closed again overnight on July 11-12. In the early hours of July 14 local time, what the UAE defense ministry described as Iranian cruise missiles struck two ADNOC-linked VLCCs in the strait’s southern lane, inside Omani territorial waters per the same statement; the IMO’s incident listing records two seafarer deaths and three injured aboard the Al Bahyah and eleven injured on the Mombasa B (the UAE ministry’s initial statement had placed a single fatality on the Mombasa, and three Al Bahyah crew were still reported missing as of July 15). CENTCOM resumed the blockade at 4:00 p.m. ET on July 14, stating it targets only vessels to and from Iranian ports, and not neutral transit.

The four days after our first edition compressed the whole pattern. India’s shipping regulator barred Indian seafarers from Hormuz-transiting voyages on July 16, citing the week’s casualties. General License X1, the last US sanctions waiver for Iranian oil, expired at 12:01 a.m. eastern time on July 17, ending the only publicly documented sanctions relief implemented under the MOU (OFAC’s own documents never name the memorandum; the link rests on timing and reporting). The IRGC claimed that two tankers had struck mines south of the strait on July 17-18; CENTCOM’s reply, in full: “Like most IRGC claims, this is false.” The three Al Bahyah crew reported missing on July 15 remain unaccounted for in any record we can locate.

July 18 widened the map and closed the file on the MOU. Kuwait Petroleum Corporation said one of its oil facilities was hit that Saturday by “repeated Iranian attacks,” with significant material damage and some injuries, and Kuwaiti authorities separately reported Iranian strikes on two power and water desalination plants, the war’s clearest spillover into third-country energy infrastructure. Deputy foreign minister Kazem Gharibabadi said the same day that Iran had “suspended our commitments under the Islamabad memorandum of understanding”; the remarks were spoken, carried by state media, and no formal instrument has surfaced on either side. Khamenei’s office called Trump’s signature “utterly worthless,” and Trump, in a NewsNation phone interview the same day, answered: “I couldn’t care less.” The IRGC claimed “full control” of the strait on July 19, saying four vessels had defied its instructions, two suffering “accidents” and two turning back. CENTCOM’s blockade ledger stood at six commercial vessels redirected and one disabled as of July 19; the disabled ship is the Curacao-flagged Belma, stopped on July 15 by Hellfire missiles fired into its smokestack.

Across the war’s first twenty weeks Iran issued six formal closure or re-closure declarations by our ledger, five under a stricter rule. The rule: a formal public statement by an Iranian state or military organ declaring the strait closed or re-closed, which excludes the February 28 radio enforcement warnings, threat statements, and “full control” assertions. The dated ledger: March 2 (Jabari); March 27, the marginal item, closing the strait to US- and Israel-linked shipping while simultaneously opening a tolled passage for others (per USNI News reporting; counting it as a fresh declaration is what separates six from five); April 18, answering the US blockade; June 11, closing to all tankers under threat of fire; June 20, three days after the MOU, citing Israeli strikes in Lebanon; and the July 11-12 re-closure. The IRGC’s July 19 “full control” statement reads as enforcement of the July 11-12 closure rather than a seventh declaration, and the ledger stays at six. The cadence is the finding: each declaration took effect in hours.

One more July event completes the pattern. On July 13 President Trump announced the US would become “guardian” of the strait and collect a 20% toll on transiting cargo; the IMO rejected the idea the same day for want of any legal basis, and the plan was rescinded within roughly 24 hours. Iran’s own tolling had the longer run: vessels were reported paying for safe or escorted passage (up to $2 million per transit, in yuan, per reporting we could not verify in aggregate) from the war’s early weeks, and the arrangement was institutionalized on May 18 as the Persian Gulf Strait Authority. The authority collected through every ceasefire until July 12, when it announced that transit through the strait was suspended, citing US military activity; it has not resumed collecting by any account located.

Figure 1. The same strait fought three different wars: Iranian exports never stopped until the July blockade constrained them, tankers collapsed and then ran escorted, and LNG went to zero for 71 days and again after the July 11 exit. Timeline schematic; segment boundaries per the sources cited in this section. Chart: Voyager Research; data from EIA/Kpler, IEA, IMF PortWatch, and wire reports, February 28 to July 20, 2026. Functional closure by cargo class and time-window, re-imposed at will; the durable seal stayed out of reach.

The closure that did happen

The clean test of closure is the cargo class with no alternatives, and there the answer was yes, for 71 days. The EIA reports, citing Kpler tracking data and hedged accordingly, that no laden LNG vessel is known to have crossed the strait between March 1 and April 24; the first laden transit after the war began came only on May 10, and that single cargo was an Iran-approved, government-to-government delivery to Pakistan, with routine Qatari transits still suspended. The sharpest illustration is the attempt that failed: on April 6 two loaded Qatari carriers, Al Daayen and Rasheeda, sailed for the strait and turned back after IRGC warnings. Loaded ships waited; the channel stayed shut.

The second LNG closure is running now. The last laden carrier through was ADNOC’s Al Hamra on July 11, per Kpler, whose July 15 snapshot found no LNG carrier entering or exiting the Gulf since; S&P Global’s count had the 10-day moving average at 0.2 laden transits a day by July 15, with one laden vessel apparently attempting a transit on the morning of July 17 whose outcome no tracker has published. Twenty-one LNG carriers were sitting inside the Gulf mid-week, roughly 1.9 million tonnes of capacity bottled up. We can confirm the zero cadence only through July 17; the closing days of this window rest on the IRGC’s own statements, and we do not adjudicate them.

The LNG closure had a supply-side twin. Iranian strikes halted Ras Laffan production in early March, and the March 18 missile strike damaged two LNG trains and a gas-to-liquids unit; QatarEnergy’s CEO has put the cumulative damage at 12.8 million tonnes a year sidelined for three to five years, roughly 17% of Qatar’s export capacity, near $20 billion a year in revenue. QatarEnergy declared force majeure on part of its long-term contract book in late March and has extended it into a fourth month, with 21 cargoes cancelled through September (unchanged as of its last confirmation, July 1). A June 21 explosion during a restart at the Barzan facility, an accident by the government’s account, killed at least 13 workers and pushed recovery further out. The IEA counts Gulf LNG output down 35 billion cubic meters year-on-year for March through June. The EIA sized the disruption at over 10 billion cubic feet a day, approximately 20% of global LNG supply.

Oil kept leaking through, which is the other half of the thesis. By March 12 the IEA had strait flows below 10% of pre-crisis levels; the EIA’s quarterly series shows Q1-2026 at 14.6 million barrels a day against 20.7 the prior quarter, the quarterly average masking a far deeper March trough. On the PortWatch spine, March was the deepest month: 4.8 transits a day, 95% below baseline, with two zero-transit days on March 13-14; April averaged 8.5 a day, and by Kpler’s own count 191 vessels crossed in all of April, around 6% of that tracker’s ~3,000-a-month baseline. June recovered to 16.1 million barrels a day of total Gulf exports on the IEA’s bypass-inclusive measure. Then July re-broke it: PortWatch logged 10 transits on July 12 against its ~90-a-day baseline, and the public series has not updated since; Kpler-derived counts, reaching us through republished reporting we label as such, put July 16 at 8 ships, a three-week low.

The machinery of denial

The tools are cheap relative to what they suppress. Mines first: a US intelligence count put “at least a dozen” in the water by March 23; a leaked congressional briefing put the count above 20 by April 22; the IMO’s June 26 estimate had roughly 80 in the historic shipping lanes, still the last official figure as of July 20. Those are three different estimators at three dates; we treat the arc as directional only, and the direction is unambiguous. Most of the growth happened while ceasefires nominally held. Against the DIA’s 2019 estimate of a 5,000-plus mine inventory, Iran has used well under 5% of its stock. The restraint reading cuts both ways, and we return to it below.

Missiles and drones did the visible work: tankers, a container ship set ablaze on July 11, the July 14 VLCC strikes, and a chemical tanker hit roughly 40 nautical miles off Oman’s coast at Qalhat, well outside the strait, in the same wave. The threat envelope also extends to the chokepoint’s alternatives: drone strikes cut 700,000 barrels a day off Saudi Arabia’s East-West pipeline for about four days (struck April 8, hours after the ceasefire announcement, restored April 12; the IRGC claimed the strike, and the Saudi restoration statement did not attribute it), the May 4 Fujairah storage fires are attributed to Iran by the UAE and denied by Tehran, and the July 18 Kuwait strikes carried the pattern to a third country’s plants.

Against this, escorted and assisted transits moved real volume on peak days, and the ships that did transit skewed large. CENTCOM claims more than 800 commercial vessels moved with US facilitation since early May, a belligerent figure we cite as such. The escorts ran, and denial priced them: every escorted barrel carried the war-risk premium, the delay, and the crew hazard that the next section quantifies. The blockade now runs the other way too: the redirects and the Belma disabling are the first sustained enforcement against Iran-bound traffic since April.

The reopening problem

Reopening runs on three clocks: certification, the actuarial table, and nerve.

Certification first. The June 17 MOU’s fifth paragraph says de-mining by Iran “will be instated within 30 days,” language that requires only a start. The window closed on July 17 with no start we can find: no clearance activity appears in any account located. The next day brought a fresh Iranian mining claim, denied by CENTCOM, followed by the suspension statement. The multinational clearance mission Europe assembled has not been heard from since a setback report of July 6-7. Iran’s deputy foreign minister insists de-mining is solely Iran’s to perform; US officials claim, attributed and unverifiable, that Iran cannot fully map its own minefields. Clearance estimates run from 40-50 days to about two months (Italy’s mine-countermeasures commander, per the coverage above), with a leaked six-month figure the Pentagon disputed the next day. Whatever the true number, no party can currently certify the lanes, and an uncertified lane holds a premium regardless of ceasefires.

The actuarial clock is the one the war has taught us the most about. Immediately pre-war, Gulf war-risk ran 0.15-0.25% of hull value per transit (from a deep-peacetime convention nearer 0.10-0.125%). By March 11, Hormuz-transit quotes ran 2.5-5% of hull value, with US-, UK-, and Israel-linked tonnage quoted at 10% or more: ten to fourteen million dollars to move one $138 million VLCC once. A separately-scoped weekly rate peaked near 2.5% and eased to about 1% by late March as more ships crossed and talks progressed; the post-MOU market settled around 2%, eight times the immediate pre-war top. The July re-escalation has kept repricing: Marsh’s global marine head put Hormuz cover at 3-10% of hull value on July 17 (a single-outlet quote we have not seen corroborated), against the 2-6% he cited on July 9; a separate July 10 market estimate called 5% the emerging norm, and the LMA’s Neil Roberts would commit only to “variable, given the continuing volatility.” The Joint War Committee has issued no listed-areas circular since March 3’s JWLA-033, an evidence gap rather than a no-change. Two corrections to the popular narrative belong here, because the insurance system was blamed for more than it did. The early-March “cover withdrawal” was a 72-hour, reinsurer-driven cancellation cascade that hit the fixed-premium, non-mutual books (fixed P&I, charterers’ liability, specialist covers) across at least seven International Group clubs; Lloyd’s List documented that mutual P&I war cover was never cancelled, and the LMA’s March 23 market statement reported that 88% of surveyed Lloyd’s war underwriters retained appetite: safety, not insurance availability, moved the traffic. Through every phase, cover remained purchasable at a price, and traffic volumes tracked that price.

Figure 3. The actuarial clock: quotes for one Hormuz transit as a share of hull value, by phase, through July 17, 2026. Ranges as reported; bases differ by product and are stated in the text; the July 17 range is a single broker’s quote. Chart: Voyager Research; data from Lloyd’s List, S&P Global, The National, and market reporting, 2026.

The slowest clock is confidence. QatarEnergy has told buyers it could reach half of output within a month of a genuine reopening and 80% a month later, with full restoration in years. Shipowners re-rate faster, though not instantly; charterers book around the strait when they can. India’s July 16 seafarer ban adds a labor constraint that outlasts any single ceasefire. The bypass that exists is real and bounded: the IEA’s March report carries the pipeline capacities, and analyst syntheses of those figures put combined additional Saudi-UAE egress near 3.5-5.5 million barrels a day against roughly 20 million of strait flow. Saudi Arabia has run East-West at its 7 million capacity and is studying a multi-year expansion; the UAE’s second Fujairah line is half-built with a 2027 target; none of it moves a single LNG cargo. The bypass also inherits the next chokepoint: rerouted Saudi barrels exit into the Red Sea, which trade press now calls the world’s most vulnerable shipping lane, and the Kuwait strikes put the region’s landside plant on the target list too.

Figure 4. The bypass arithmetic: pipeline alternatives cap out near a quarter of pre-war strait flow, and none of them carries LNG. Chart: Voyager Research; data from IEA March 2026 Oil Market Report (capacities) and analyst syntheses (combined egress estimate).

Q: Did Iran close the strait?

Pick the metric before answering, because the metrics disagree by design. By vessel count, the worst windows ran at roughly 2-10% of baseline: severe suppression, with some ships still moving. By volume, the floor was higher, since the ships that moved were disproportionately supertankers under escort. By cargo class, LNG was fully closed for 71 days, a genuine closure of one fifth of the world’s LNG supply, and a second LNG closure has run since the July 11 exit. And by Iran’s own traffic, the strait stayed open throughout the truce months: its exports ran in every phase, including 12 million barrels moved dark in the single week between the July waiver revocation and the blockade’s return.

So “closure” was never the right noun. Iran demonstrated functional closure by cargo class and time-window: total for LNG for ten weeks and again now, near-total for hull traffic for stretches of March and April, partial and priced everywhere else, re-imposable in hours each time a truce broke. A full seal never appeared, and Iran’s own shipping behavior suggests it was never the goal; a sealed strait would trap Iranian barrels too. The July blockade adds the mirror image: for the first time since April, Washington is enforcing the constraint on Iranian barrels.

Q: Can Iran afford the strategy?

The pre-war intuition said no: a country that ships ~90% of its oil from Kharg Island, inside the Gulf, with a bypass at Jask that moved under 70,000 barrels a day pre-war against a $2 billion, never-finished design, cannot afford to burn the waterway it lives on. The observed war complicates that arithmetic. In the tension month before the April blockade, Iran’s oil revenue ran above its early-February run-rate on both major tracker bases, up roughly 44% on the Kpler-derived calculation (Al Jazeera’s own multiplication of tracked liftings by a conservative $90 floor) and up modestly on the more cautious basis; prices spiked while Iran’s volumes held at or above trend. After the June MOU it shipped more than 80 million barrels in 26 days by the highest tally (secondary tallies run 60 to 80 million; Iran’s own negotiator claimed 40 million-plus with a price premium). It monetized the chokepoint directly through transit fees. On the truce-era evidence through June, disruption ran revenue-positive for Iran.

The July window sharpened the cost side. The PGSA suspended its own tolled passage on July 12, and the July 17 waiver expiry ended the last lawful dollar-cleared wind-down of Iranian oil transactions. TankerTrackers estimated on July 14 that roughly 50 million barrels of truce-era liftings would sit stranded if the blockade held, a conditional projection we label as such, and nine sanctioned tankers went dark off Malaysia carrying roughly $989 million of crude bound for Shandong. The costs also arrive through the customer. China took 97.6% of the Iranian oil afloat when the April blockade hit; by June, Chinese imports of Iranian crude had halved to roughly 654,000 barrels a day, Shandong’s independent refiners were buying Qatari, Iraqi, and Emirati barrels in size, and Iranian discounts were narrowing toward uncompetitive. A seller whose single indispensable customer is learning to substitute is spending its franchise, and what was revenue-positive in the truce months now runs through a blockade with the waiver gone and the toll suspended. We still cannot cleanly attribute the export collapse to Iranian choices: the US blockade imposed much of it, so the data cannot separate “cannot afford” from “was prevented.” The question stays open, and the Soundings below carry it.

Q: What would reopening actually take?

A settlement that survives its own window comes first, and there is no window in practice: the 60-day clock nominally runs to August 16, Iran said on July 18 it has suspended its commitments, and neither capital still mentions the date. Mine clearance somebody can certify comes next, against Iran’s insistence that only it may clear and the attributed US claim that it cannot fully map its own fields. Insurance re-rating follows, and the 2026 record suggests it is the fast part, weeks once risk genuinely recedes, because appetite never left. Cargo-class recovery then runs on its own physics: tankers return in weeks; Qatari LNG needs months to years, because the damage is onshore as much as maritime. At every step, one missile resets the clock. Every reopening clock runs in months; re-closure has twice taken a day.

What the market prices

The war’s price signature is itself evidence for the thesis. EIA’s settlement-grade Brent spot peaked at $138.21 on April 7, the day Dated Brent printed $144.42, the highest physical-benchmark assessment since the series began in 1987.

Fourteen weeks later the re-escalation repriced in stages, and we read it on the futures spine, the only series retrievable at exchange grade this run. ICE Brent settled at $83.30 on July 13, $84.73 on July 14, and $88.10 on July 17, a 5.8% week that priced the VLCC strikes and the blockade’s return. A thin weekend session then carried the suspension and Kuwait news to $91.41, and the entire spike was flat by mid-morning London on July 20. (The EIA spot series our July 16 edition quoted was not retrievable live this run; its July 13 print was $81.62. The spot and futures spines differ, and we do not mix them.)

Figure 2. The repricing: April’s all-time records against July’s second look. April marks are spot and assessment grade; July marks are ICE futures settles, a different spine, labeled. Chart: Voyager Research; data from EIA and S&P Global (April), ICE via market reporting (July), 2026. The four days test the July 16 reading, and the reading bent without breaking. The market watched the MOU die and the war reach Kuwait’s plants, then moved the benchmark under six percent and faded its own panic spike within hours; it is still pricing a chronic tax on transit. A certified mine strike on an escorted VLCC would still reprice everything, and nothing in the window tested it; the Soundings below watch both branches.

Soundings

Where we will measure, and what depth changes the Assessment:

1. ~August 16, 2026. The 60-day window’s nominal end. A lapse that passes unremarked entrenches the cycle; any revival that survives past the date reopens the reopening sequence.
2. The LNG streak. The first tracker-grade print covering July 18-20 (Kpler, S&P Global), the outcome of the July 17 attempted transit, and QatarEnergy’s force majeure against its September horizon. A completed laden transit would be the first crack in the second cargo-class closure.
3. The mine ledger. The next official count after the IMO’s ~80 of June 26, any certified cleared lane, and whether the European clearance mission surfaces at all.
4. The war-risk quote. A second broker source on the July 17 range of 3-10%, and any Joint War Committee circular after March 3. A durable return below ~1% per transit would count against the thesis.
5. PortWatch. Whether the public series resumes past its July 12 stall; thirty consecutive days above 60 transits remains our bar for durable normalization.
6. The PGSA toll. Whether Iran resumes collecting when passage resumes. Tolling restored would re-arm the monetization finding; a dead authority would date the strategy’s revenue phase as truce-only.
7. The Kuwait front. Whether strikes on third-country energy infrastructure recur. July 18’s strikes drove the weekend spike that faded; a pattern that holds would move the market question from transit tax to regional supply.

Voyager Research · Independent analytical research. Not personalized investment advice. · Sources are cited inline; tanker-tracker figures are labeled and carry their providers’ scope caveats; belligerent claims are attributed, and we do not adjudicate them. · Data anchor July 20, 2026. This edition updates the July 16, 2026 edition and supersedes its data anchor; The Log grades that edition’s Soundings. Corrections are appended, never silently edited.