Strait of Hormuz Transits Collapse 77% to Just Five Vessels in a Single Day

The Analysis

Spread the map out on the table and put your finger on the head of the Persian Gulf, then drag it southeast until the water opens into the Gulf of Oman and, beyond that, the Indian Ocean. Everything that moves by sea between those two points passes through a single gap, and that gap is not a preference, not a commercial choice, not something a chartering desk can negotiate its way around. It is a door. For something close to a fifth of the world's seaborne oil it is the only door in a wall that runs from the mountains on one side to the desert on the other, and no amount of planning makes that wall any shorter.

On 30 July 2026 the vessel-tracking data counted five ships going through that door. The day before, the same count was twenty-two. That is a 77 percent collapse between one day and the next, and every one of the five ships that did move used the channel running along the Iranian side. On the same day the analysts at Kpler said out loud what the arithmetic already implied: the thing holding Gulf crude exports back is not berth space, not loading windows, not a shortage of buyers. It is the security risk of putting a hull and a crew into that water, and the price of insuring against it.

Now look at the other end of the same peninsula. On that same date, the Bab el-Mandeb strait recorded forty-seven transits, twenty-seven of them heading into the Red Sea and twenty coming out. Two chokepoints, one calendar day, two completely different readings, and the temptation is to average them into a single mood about the region. I want to argue against that temptation, because the difference between the two numbers is where the actual work is.

Start with what five ships does not tell you. Five is not zero. The strait did not close, no authority announced a closure, and the five hulls that moved did move through the Iranian-side channel rather than turning back. If your mental model of a chokepoint risk is a light switch that flips from open to shut, this data point does not belong to it. What you are looking at is more like a valve being turned down, and the interesting questions are how far down, for how long, and what price the remaining flow commands.

Two days of data is not a baseline either. I have twenty-two and I have five, and I do not have a verified figure for what an ordinary day looks like on that route, so I am not going to invent one and neither should you. If you want a threshold you can manage against, you have to build it yourself, off the same kind of tracking data, over at least thirty consecutive days, and you have to write down the date you built it on because it will go stale. A trigger that fires on a single day's print will have you rerouting cargo every week and paying for the privilege.

What the Kpler note adds is a mechanism. A drop in transits can be demand, it can be maintenance, it can be a scheduling artefact of how a particular counting method treats a ship that loiters. When the analysts on the same day point at persistent maritime security risk and at war-risk insurance costs as the binding constraint on Gulf crude exports, they are telling you the valve is being turned by underwriting, not by loading programmes. That distinction changes what you do about it. You cannot negotiate with geography, but you can decide how much exposure you carry, how long you carry it, and what you pay to lay it off.

Now put the map back on the table and look at it on the scale I prefer, which is not days but decades. The water between Iran and the Arabian Peninsula has been narrow for as long as anyone has been shipping anything through it. Ten years ago this stretch looked like a different place, and not because the coastline moved. What has changed is how much of the world's industrial metabolism is routed through a piece of water you can see across on a clear day, and how tightly the scheduling of that traffic is coupled to everything downstream of it. A refinery in Asia, a polymer plant fed by its output, a converter making packaging, a retailer with a shelf-life problem: none of them touch the strait, all of them sit downstream of it.

That is the part I keep coming back to. The network did not get more fragile in the abstract. It got denser, it got faster, and it got less tolerant of slack. Inventories were rationalised over two decades because holding stock was expensive and the water seemed reliable. Contract lead times were tightened because speed was a competitive weapon. Every one of those decisions was sensible on its own, and together they converted a slow-moving geographic fact into a fast-moving commercial one. When the print goes from twenty-two to five, the thing that hurts is not the strait. It is the missing weeks of cover that everyone agreed to stop paying for.

So what does a 77 percent single-day drop mean for someone whose job is moving cargo in and out of this region? Let me split it, because importers and exporters sit on opposite sides of the same door.

If you buy Gulf-origin material, your exposure is on the inbound leg and it shows up first as lead time, not as price. Crude, LPG, methanol, polymers, sulphur, aluminium, fertilisers, bitumen: the categories differ but the mechanics are identical. Your contract assumes a loading window and a transit time, your production plan assumes a delivery date derived from those, and your inventory policy assumes the distribution of arrivals you have historically observed. A single day at five transits does not break any of that. A fortnight at that level rearranges all of it. The sequence is always the same: quote validity shortens first, then nominations slip, then transit time stretches, then your safety stock is doing a job it was never sized for.

If you sell into the Gulf, your exposure is outbound and it shows up differently. Your problem is not getting your own cargo loaded, it is whether the vessels, the equipment and the insurance appetite exist to take it there at a price your buyer will accept. Inbound capacity and empty equipment positioning are the two things that quietly degrade when a lane becomes expensive to underwrite. Containers that would have been repositioned into the Gulf sit where they are. Boxes that would have been pulled out stay longer. The symptom at your end is not a rate, it is a booking that is hard to confirm and a transit that is quoted with more caveats than last quarter. Watch bookings and equipment availability as your early indicators, not headlines.

Let me put numbers on the first case, because vague worry is not a plan. Take the two prints I actually have: twenty-two transits, then five. Five divided by twenty-two is 22.7 percent, which means the flow you observed on the second day was a little under a quarter of the day before, and the gap between them is the 77 percent everyone quoted. Assume, and this is my assumption not a fact, that 35 percent of a particular feedstock you consume comes from Gulf-origin suppliers, that your normal door-to-door transit is eighteen days, and that you hold twenty-one days of cover against it. Assume also that the reduced flow persists for two weeks rather than a single day, because a single day is noise and two weeks is a disruption.

Under those assumptions, the Gulf share of your supply is arriving at roughly 22.7 percent of its usual rate. In a normal week, the Gulf leg covers about two and a half days of your consumption, which is seven days multiplied by your 35 percent share. At a quarter of the usual rate it covers about half a day. The weekly shortfall is therefore close to two days of consumption, and twenty-one days of cover is consumed in something like eleven weeks. Extend the disruption assumption to a month and the arithmetic gets uglier in a hurry. Raise your cover from twenty-one days to thirty and you buy yourself roughly five extra weeks of runway under the same shortfall. That is the whole argument for a few more days of stock, expressed as weeks instead of as a feeling.

Now the money, and I am going to be careful here because I do not have a verified number for what war-risk cover costs and I am not going to print one I cannot source. Call the loading you pay per shipment V, expressed as a fraction of cargo value, and call your monthly cargo value M. Your monthly war-risk cost is M times V, and your annual figure is twelve M V. If that loading were to go to four times its current level, the incremental annual cost is thirty-six M V. Against that, holding an extra fourteen days of inventory ties up roughly half a month of cargo value, which at an annual carrying cost of 20 percent, another assumption you should replace with your own finance team's number, costs you about 9 percent of one month's cargo value per year.

Set those two against each other and you get a decision rule that survives not knowing V. Thirty-six M V equals that inventory cost when V is about 0.26 percent of cargo value per shipment. Below that, a fourfold jump in the war-risk loading is cheaper than carrying a fortnight of extra stock. Above it, the stock wins. You do not need a market quote to use this, you need your own last invoice and a calculator, and the answer will move as your own numbers move. That is the point of doing the arithmetic with a symbol in it.

The second thing worth pricing is the alternative, and here the map is merciless. If your cargo loads inside the Gulf, the forty-seven transits through Bab el-Mandeb are irrelevant to you. That strait sits at the far end of the peninsula and governs access to the Red Sea and the Suez routing beyond it. A busy Bab el-Mandeb does not create a second exit from the Persian Gulf, because there isn't one. I have watched otherwise careful teams comfort themselves with exactly this confusion: activity somewhere in the region read as capacity everywhere in the region. It is not the same water.

There are overland routes out of the Gulf that the industry talks about by name, and I am deliberately not going to attach capacity figures to them because I have not verified those figures against anything I can point you to. If they matter to your supply plan, get the numbers from your supplier or your own reference material and treat them as what they are: partial, conditional and subject to their own politics. The useful observation is structural rather than numerical. Those corridors mostly serve bulk liquid energy, not the containerised or breakbulk cargo most of this readership moves, so for a great many importers the honest answer is that there is no physical bypass at all, only a delay and a premium.

For cargo that originates outside the Gulf, the routing conversation is completely different and worth having separately. A longer sea route around the southern tip of Africa is a real option when the origin and the destination let you choose, and the cost of it is time and distance, not a strait. But note what that option does not do: it does not help a molecule that has to leave Ras Tanura or Jebel Ali or any other terminal inside the water. Conflating the two situations is the single most expensive analytical error on this topic, because it leads people to buy optionality they can never exercise.

So here is what I would actually do with this week's data, and I want to be specific about dates and owners rather than gestures.

By the middle of August, the person who owns your inbound planning should be able to hand you a single page that lists every input with Gulf-origin content above ten percent, ranked by days of cover rather than by spend. Spend rankings hide the problem, because the cheap input with nine days of cover is the one that stops your line. If that page does not exist yet, building it is the highest-value thing on the list and it is two days of work with a spreadsheet and a purchasing contact.

Inside ten working days, re-quote the war-risk loading on your three largest Gulf lanes and insist the quote carries a written validity window. A quote valid for a month is not a quote in this environment, it is a wish. Seventy-two hours is a reasonable ask. Set your own re-quote trigger at a twenty-five percent move from the last booked voyage, and write the trigger down, because a rule you have not written down is a rule you will argue with under pressure.

Before the end of the third quarter, move days-of-cover on Gulf-origin feedstock from wherever it sits today to a floor of thirty days, and put a trigger on the increase rather than doing it once: seven-day rolling average of daily transits below fifteen, sustained for five consecutive days. The number fifteen is not from the data, it is a planning placeholder roughly halfway between the two prints I do have, and you should replace it the moment your own thirty-day baseline exists.

By the end of October, have at least one qualified non-Gulf source for each of your top three Gulf-origin inputs, qualified in the sense that you have bought from them and run the material, not in the sense that someone has a business card. Target twenty percent of volume on those inputs by the middle of next year. Qualification lead time is the thing that eats you in a disruption, because a source you have never bought from is not a source, it is a phone number.

And set the review cadence now, before the news cycle moves on. Re-draw the network map quarterly, not annually, and set a hard target that no single chokepoint carries more than forty percent of any critical input by the end of 2027. That is a multi-year target on purpose. You cannot fix a decade of concentration in a quarter, and pretending otherwise is how plans fail.

The traps are worth naming because they are all versions of the same mistake, which is reading a number as more than it is. A single-day print is not a trend, and a two-day dataset is not a baseline. A busy neighbouring strait is not a bypass for a route it does not serve. A quote without a validity window is not a price. An alternative source you have never bought from is not an alternative. A supplier concentration that sits at tier two or tier three of your bill of materials does not show up in your spend analysis, and it is frequently where the real exposure hides, because your supplier's supplier is standing in the same queue at the same door.

There is also the opposite error, which is assuming that because the strait stayed open, nothing happened. Twenty-two transits and five transits are both snapshots, and neither of them tells you what a normal day looks like. What you can say, on the evidence in front of you, is that a route carrying something like a fifth of the world's seaborne oil printed a day with five movements on it, and that the people who watch that water for a living attributed the constraint to security risk and insurance rather than to cargo. That is enough to act on, and it is not enough to panic about. The discipline is to size the exposure in days of cover and in percentage of a single input, and then to buy down whichever of those is cheapest.

There is one more version of the mistake, and it runs the other way. Some teams read five ships and immediately buy everything on offer: extra cover, extra inventory, extra optionality, all of it priced at the worst moment in the cycle. Optionality is not free. Cover you buy after the event is priced after the event, and inventory you build in a hurry is the inventory you will be discounting when the flow comes back. The discipline I am describing is the opposite of a scramble: size the exposure once, calmly, in days and in percentage of a single input, then buy down the cheapest of the available levers and stop. Done well, this looks boring from the outside. It is a spreadsheet, a written trigger, a calendar entry, and a supplier you qualified last year. Done badly it looks like activity, and it costs about the same.

Let me end where I started, with the map. The water there has looked roughly like this for a very long time, long before there was anything on it worth insuring. Ships will keep going through it, and the count will keep moving up and down, and most days the number will not be interesting. What has changed over the working life of everyone reading this is not the strait. It is the amount of the world that now sits downstream of it, and the thinness of the buffer we collectively decided to run. Ten years ago this lane was a line on a chart. Today it is a variable in your production plan, and the only real defence against a variable you cannot control is knowing exactly how many days it takes to reach you.

↑ Back to the key points

— By Tomás Ferreira