The Strait of Hormuz wasn’t closed by a naval blockade. There are no mines in the water, no Iranian warships forming a line across the channel. The waterway is open, technically. Ships can physically pass.
They just can’t get insurance.
Within 72 hours of the first US-Israeli strikes on Iran, seven of the twelve International Group P&I clubs — the institutions that collectively insure roughly 90% of the world’s ocean-going tonnage — executed identical cancellation notices for war-risk coverage across the Persian Gulf. Gard. NorthStandard. Skuld. Steamship Mutual. The American Club. One by one, the paper underpinning global oil trade was withdrawn. Not because Iranian warships were blocking the channel, but because Iran had struck eight or nine vessels with drones cheap enough to buy in bulk, and the London reinsurance market had done the actuarial math.
A Shahed-136 drone costs somewhere between $20,000 and $50,000. A PAC-3 interceptor costs $4 million. A THAAD round costs $12.7 million. For every dollar Iran spends attacking a tanker, the defender spends roughly $100 intercepting it. The math is unsustainable for the defender. The London market understood this faster than the geopolitical analysts did.
Iran closed the Strait actuarially. The question now is whether anyone understands what that means for reopening it.
The Consensus Isn’t Wrong. It’s Using the Wrong Map.
The mainstream narrative right now is not irrational. It is pattern-matching to the best available data, which is 50 years of prior oil shocks, and it is doing so in good faith.
Here is the consensus, given its due: In 1973, the Arab oil embargo quadrupled prices and rattled the global economy — but it resolved. In 1979, the Iranian Revolution disrupted roughly 4% of global supply and prices doubled — but they stabilized, eventually. In 1990, Iraq’s invasion of Kuwait spiked prices sharply — but Saudi Arabia ramped up production to compensate, strategic reserves bridged the gap, and the crisis contained itself. In every prior instance, the disruption was painful, temporary, and bounded by the existence of producers who could absorb it.
The consensus is applying that pattern to what it sees today: a conflict that Trump says will last 4-5 weeks, a Strategic Petroleum Reserve holding 415 million barrels, pipeline bypass routes in Saudi Arabia and the UAE, and Chinese floating storage that was unusually full when the conflict began. Layer in demand destruction — high prices suppress consumption — and the model outputs: volatility, not sustained shock. Brent briefly tests higher, then normalizes toward $70 by year-end.
This is not a stupid view. It is the view that every serious oil shock in history supports.
The problem is that this crisis has three structural features that none of those historical shocks share. And the consensus model hasn’t priced any of them.
How It Actually Closed
Before the arithmetic, the mechanism matters, because the mechanism determines the timeline for reopening.
In every prior Gulf disruption — the Tanker War of 1987-88, the Gulf War of 1990-91, the various near-closure threats of the 2000s and 2010s — the threat to shipping was kinetic. Mines, anti-ship missiles, naval vessels. The closure mechanism and the reopening mechanism operated through the same domain: military force. You could escort convoys, sweep mines, destroy missile batteries, and eventually restore passage. The problem and the solution were the same kind of thing.
This time they aren’t.
Iran didn’t deploy its navy. It deployed cheap drones and cheaper drone boats — unmanned surface vessels — in enough volume to make insurers do the math. Eight or nine vessels struck in ten days. One Honduran-flagged tanker hit by two aerial drones and set on fire in the Strait on March 2. One Marshall Islands-flagged vessel struck by a kamikaze drone boat off Oman’s coast — reportedly the first successful unmanned surface vessel strike of the war, one crew member killed. After the Prima ignored IRGC warnings and was struck by drone on March 7, major carriers stopped sending ships entirely. Not because passage was physically impossible. Because it was uninsurable.
The reinsurance cascade that followed was institutional, not military. London treaty reinsurers determined they could no longer satisfy the 99.5% Value-at-Risk capital charges required under EU Solvency II rules. When they pulled back, the P&I clubs had no choice. When the clubs pulled back, shipowners had no coverage. When shipowners had no coverage, they stopped sailing. The whole sequence took 72 hours.
The implications for reopening are severe. A ceasefire doesn’t restore the insurance market. Destroying every Iranian drone launcher doesn’t restore the insurance market. The reopening requires reinsurers to rebuild risk models incorporating new conflict data, individual voyages to be re-underwritten across the entire chain, and risk committees at major energy companies to sign off on direct transits. These are sequential institutional processes. They cannot be compressed by executive order, naval presence, or presidential tweets.
The Red Sea is the calibration anchor. The 2023-2025 Houthi disruption produced 26 months of insurance-driven rerouting despite continuous military operations. Premiums never returned to baseline. Kpler’s assessment: the path back to normalcy is measured in quarters, not weeks.
The Trump administration announced a $20 billion DFC reinsurance backstop on March 6. As of March 9, it has produced no confirmed insured VLCC transits at scale. JPMorgan estimates total insurance exposure for vessels in the Gulf at $352 billion — nearly twice the DFC’s $205 billion statutory ceiling.
The Strait closed in hours. It will not reopen on the same timescale.
The Supply Math No One Is Running
Now the arithmetic.
The historical analogs the consensus is citing are the wrong size for this disruption. The 1973 embargo disrupted roughly 7% of global oil supply. The 1979 Iranian Revolution disrupted approximately 4% — and that 4% disruption doubled prices over 12 months. The 1990 Gulf War disrupted the production of two countries: Iraq and Kuwait, roughly 4.3 million barrels per day.
The Hormuz closure has disrupted approximately 20 million barrels per day — the effective throughput of the Strait. It has done this not by shutting down one or two producers, but by bottling up the exit point for seven of them simultaneously. Saudi Arabia. UAE. Iraq. Kuwait. Qatar. Oman. Iran. All of them locked behind the same 55-kilometer chokepoint.
This is the structural break that no historical analog captures. In 1973, Saudi Arabia was the swing producer that absorbed the shock. In 1990, Saudi Arabia ramped production to compensate for Iraq and Kuwait. In every prior crisis, there was a major producer standing outside the disruption, able to step in.
This time, Saudi Arabia is inside the bottle.
The bypass pipelines exist, and the consensus cites them. Saudi Arabia’s East-West Pipeline can move roughly 5 million barrels per day to Red Sea terminals. The UAE’s Fujairah pipeline adds another 1.5 million barrels per day. Combined, under ideal operating conditions: perhaps 6-7 million barrels per day of alternate routing.
Against a 20 million barrel per day disruption, that’s roughly 30% coverage. The remaining 14 million barrels per day have nowhere to go.
Which brings us to the storage clock.
The Countdown That Has Already Started
JPMorgan’s commodities team published the most precise available model of what happens next. The Gulf’s seven major producers hold approximately 343 million barrels of combined onshore crude storage. Another 60 empty tankers anchored in Gulf waters can absorb roughly 50 million more. Total buffer: approximately 393 million barrels, or about 25-26 days of stranded production before storage constraints force involuntary shutdowns.
The cascade isn’t a cliff — it’s a schedule. Iraq had roughly three days of storage buffer when the conflict began. By Day 8, Iraqi southern production had already collapsed 70%, from 4.3 million barrels per day to 1.3 million barrels per day. The Rumaila field — Iraq’s largest, responsible for roughly a third of national output — halted 100% of production from its southern section. The reason was not airstrikes. The reason was that stockpiling had reached critical levels and there were no tankers to take the oil. Kuwait hits its storage limit around Day 14. Broader Saudi and UAE thresholds approach around Day 18.
As of March 9, the clock is 9 days in.
The storage model matters for a reason beyond the price implications. When a field is forced to shut in because storage is full — not because of damage, not because of deliberate production cuts, but because there is physically nowhere to put the output — the disruption changes category. It transitions from a logistics shock to upstream supply destruction. Wellbore integrity at major fields takes weeks to months to restore after uncontrolled shutdowns. Fields that were producing normally on February 28 may require significant remediation before they can return to full output, independent of when the Strait reopens.
This is the second historical break. Every prior oil shock disrupted the flow of oil. This one, past Day 25, begins destroying the production capacity itself.
The Buffer Math
The consensus response to all of the above is: strategic reserves. The IEA system holds roughly 1.2 billion barrels of public reserves plus an additional 600 million barrels under government obligation from industry. The US Strategic Petroleum Reserve alone holds 415 million barrels. A coordinated release — the G7 finance ministers were discussing 300-400 million barrels as of March 9, which would be the largest in the IEA’s 52-year history — would dwarf the 180 million barrel release during the 2022 Ukraine crisis.
This is a real tool. It would suppress prices. It is worth doing.
It does not solve the problem.
Against a 20 million barrel per day disruption, the entire IEA public reserve system covers roughly 60 days of the supply gap — assuming the gap is fully offset by reserves, which it cannot be, and assuming demand doesn’t also respond, which it will. The US SPR alone covers roughly 21 days. Even a record 400 million barrel coordinated release buys somewhere between 20 and 40 days of a full Hormuz closure, depending on assumptions.
And the insurance-driven reopening lag operates on a timeline measured in quarters.
The buffers and the reopening mechanism are not on the same timeline. The reserves run out before the institutional processes required to reopen commercial shipping are complete. That gap is what no historical analog has had to price.
For the first ten days of this conflict, the administration declined to deploy the buffer at all. Trump’s post on Truth Social on March 8: “Short term oil prices, which will drop rapidly when the destruction of the Iran nuclear threat is over, is a very small price to pay for U.S.A., and World, Safety and Peace. ONLY FOOLS WOULD THINK DIFFERENTLY.” On March 9, Energy Secretary Wright pivoted to saying the administration was “discussing coordinated releases” with other nations. The shift signals that the initial confidence is eroding. It also means the clock ran for ten days before the buffer was even engaged.
The Working Hypothesis
The market’s baseline scenario — Hormuz disruption contained within 4-6 weeks by strategic reserves and de-escalation, prices normalizing toward $70/barrel by year-end — is built on historical analogies that don’t structurally apply to this crisis.
Three things are true simultaneously that have never been true simultaneously in any prior oil shock:
First, there is no swing producer positioned outside the disruption. The primary holders of global spare capacity — Saudi Arabia, UAE — are bottled up behind the same chokepoint as everyone else. The 1973 playbook, the 1990 playbook, and every playbook in between assumed someone was standing outside the bottle. Nobody is.
Second, the closure mechanism is actuarial, not kinetic, which means military resolution does not equal commercial reopening. The Red Sea precedent suggests the reopening lag is measured in quarters. The buffer mechanisms — reserves, bypass pipelines, DFC reinsurance — cover weeks. The timelines don’t match.
Third, past Day 25, this stops being a logistics shock and starts being upstream supply destruction, as storage fills and producers are forced to shut in fields regardless of whether a ceasefire has been signed.
The falsifiable call: If the Strait remains effectively closed past March 24 — Day 25 of the conflict — the cascading forced shut-ins will begin converting this from a logistics disruption into a supply destruction event. At that point, the year-end $70 price forecast is not a baseline. It is a tail scenario. The market has not priced this. Oil at $113 is still, arguably, a market pricing in a short conflict with known historical precedents. It is not a market that has run the 25-day storage model and concluded the analogies are broken.
This is the bet. It can be lost. If the conflict resolves in three weeks and the Strait reopens faster than the Red Sea precedent suggests, the consensus was right and the map was fine.
I don’t think the map is fine.
What would change my mind
1. The DFC reinsurance backstop produces verified VLCC transits at scale within ten days, demonstrating that state-backed insurance can override the institutional processes that drove the market closure. If tankers are moving in volume by March 19, the insurance-lag thesis breaks.
2. A ceasefire emerges before March 24 and the Red Sea precedent fails to hold — insurers return to the Gulf on a timeline of weeks rather than quarters. This would require the conflict to have been sufficiently short that reinsurers can model it as a discrete, bounded event rather than an ongoing operational risk. Possible, but the Red Sea data makes it the low-probability outcome.
3. The IEA coordinated release of 300-400 million barrels successfully suppresses prices below $90 for 60 consecutive days, demonstrating that the buffers are large enough to bridge whatever reopening lag materializes. This would require both the release to be authorized quickly and the Strait to reopen within the window the reserves can cover. Both conditions would need to hold simultaneously.
If any of those three conditions are met, update accordingly. The scorecard is public.
Working Hypothesis tracks every published thesis. This call will be scored on two dimensions: direction (was the disruption longer and more severe than the consensus baseline predicted?) and calibration (was the confidence level appropriate given what was knowable on March 9?). Falsification date: March 24, 2026 for the storage threshold; June 30, 2026 for the price normalization call.