What a shipping chokepoint actually does when it fails

Article 3 of 8  ·  Series II of III  ·  Published 28 October 2026  ·  Analysis → Forecast → Recommendations

Series II, third row. Each article measures one physical layer with the same three questions — concentration, criticality, substitution time — and adds a row to one comparison table. The first two rows found dependencies with no alternatives at all. This one is different, and the difference is the most useful thing in the article.

1. The Market Knew First

In the third week of February 2026, several things happened that looked unconnected unless you were watching the right numbers.

War-risk insurance premiums for Gulf transits moved from 0.125 percent of a vessel's insured value to somewhere between 0.2 and 0.4 percent. Iran roughly tripled its oil export volumes between the fifteenth and the twentieth of the month, drawing down storage. Saudi Arabia made similar precautionary moves.

None of that was a forecast. It was behaviour. Underwriters repriced, and the parties most exposed moved their cargo while they still could.

The military escalation came on 28 February. The freight data had moved roughly a week earlier.

What followed is the subject of this article, and the useful part is not the disruption itself. It is what happened to the language used to describe it.

Some sources reported that the Strait of Hormuz was closed. Others, months later, reported that it had not closed in 2026 and that only the risk premium had become permanent. Both accounts came from competent maritime analysts with access to the same tracking data.

They were not contradicting each other. They were answering different questions.

A chokepoint does not have a switch. It has a price, a transit time and an insurance rate, and all three move continuously while the legal status stays exactly where it was.

That distinction is what this row of the table is really about, and it applies far beyond shipping.

2. What a Chokepoint Failure Actually Looks Like

The intuitive model of a chokepoint is a door. It is open or it is shut, and if it shuts, trade through it stops.

Almost nothing that has happened since late 2023 fits that model.

The Red Sea, measured

Attacks on shipping from late 2023 pushed the major container lines onto the route around the Cape of Good Hope. That added roughly 3,500 nautical miles and ten to fourteen days to the Asia-to-Europe voyage.

Two years later, the corridor has never been closed and traffic has not returned. Red Sea and Bab el-Mandeb transits in 2025 ran at roughly 35 to 40 percent of 2023 volumes. Suez Canal traffic in early 2026 was still about 60 percent below its pre-diversion level.

By late August 2026 the picture had improved without normalising. Around 290 ships were tracked through the canal's northern approach in one week, on par with the week before, which is about 36 percent below normal levels and roughly 30 percent higher than 2025. Traffic through Bab el-Mandeb was down about 15 percent.

A route that is legally open, physically passable, and running at two thirds of its former volume three years after the first attack. That is not a door. It is a dimmer.

Hormuz, and the vocabulary problem

The same pattern, more sharply. Roughly a fifth of the world's daily oil and a comparable share of its liquefied natural gas moved through the Strait of Hormuz before February 2026.

After the escalation, transits fell severely. Gulf producers rerouted: Saudi Arabia activated a pipeline carrying crude west to the Red Sea port of Yanbu, and the United Arab Emirates pushed volumes through the Fujairah bypass, which sits east of the strait and sends cargo to Asia.

Whether to call that closed depends entirely on what you are measuring. Legally, nothing was closed. Commercially, a great deal of traffic stopped using it. Both statements are accurate and they support opposite decisions.

Running the instrument

This is the first row in Series II with genuine alternatives, and the reading is dramatically better than electricity or advanced packaging. A ship can be rerouted in a week. A grid connection takes five years, and a second-source packaging line takes longer than that.

Which raises the obvious question. If chokepoints are the substrate with the best substitutability, why do they produce the most visible disruption of any of them?

The two sections that follow answer that, and the answer is not what the alarm suggests.

3. Subtheme One — The Adaptation Is the Cost

The system rerouted, quickly and competently. Nothing ran out. Shelves did not empty. That is the striking fact and it is usually reported as reassurance.

It is also where the entire cost sits.

What rerouting actually consumes

Against all of that, avoiding the canal saves its transit fees — reported at $300,000 to $700,000 per vessel — which is why the arithmetic is not always obviously against rerouting for a single voyage. The costs that dominate are the ones spread across time and capacity rather than charged as a fee.

One industry estimate puts the annual hit to global trade at fifteen to twenty billion dollars. Whatever the precise figure, it is a permanent operating overhead rather than a loss event.

Why this makes it invisible

A closure would produce shortages, and shortages produce attention. A reroute produces a slightly higher price, a slightly later delivery, and a slightly tighter shipping market.

So the successful adaptation is precisely what prevents the cost from being noticed. Nothing fails, so nothing is investigated. The system is paying a permanent premium that shows up as ordinary inflation in freight rates, and no single decision-maker ever sees a line item that says chokepoint.

This is the argument Series I made about institutions, arriving in logistics. Adaptation has not stopped. It has become more expensive, and the expense is spread thinly enough that nobody is prompted to add it up.

Who actually pays

The premium does not stop with the carrier. It moves down the chain in a specific order, and the order determines who notices.

The carrier absorbs it first, then recovers it through freight rates. The importer pays the rate, and also carries ten to fourteen extra days of goods in transit, which is working capital tied up in cargo sitting on water. A firm importing a container a week from Asia is permanently financing two additional weeks of inventory it did not finance in 2019.

From there it reaches the final buyer as a slightly higher price, mixed in with everything else, attributable to nothing.

Which is why no board paper has ever contained a line called chokepoint. The cost is real, it is continuous, and by the time it arrives anywhere a decision-maker would see it, it has been blended into freight inflation, working capital and margin — three places that each have a dozen other explanations.

And the fleet effect, which is the one people miss

Longer voyages do something subtler than raising costs. They consume ships.

The same cargo moved over a longer distance requires more vessel-days, so a fixed fleet carries less. That tightens capacity across every route, including routes nowhere near the disruption, and it raises rates for shippers who never touch the Red Sea.

It also means that a return to the canal would release capacity suddenly, putting downward pressure on rates. Which gives the industry an odd exposure: a security improvement in the Red Sea would be a commercial shock to carriers, and that is one reason analysts expect any return to be gradual rather than sharp.

4. Subtheme Two — Risk Does Not Disappear, It Normalises

The second thing this row shows is why recovery lags the security situation by years.

Attacks fell. Traffic did not return

Attack frequency in the Red Sea declined sharply through 2025. Insurance premiums came down from their peaks. And traffic in late 2026 was still around a third below normal.

That gap is not irrationality. Carriers are not pricing the number of incidents. They are pricing insurance classification, contract structure, schedule reliability and the risk of being the one vessel that is hit.

A carrier that returns to the canal and loses a ship has made a catastrophic decision. A carrier that stays on the longer route and pays more has made a merely expensive one. Those two outcomes are not symmetric, and no fall in incident frequency makes them symmetric.

So the risk was not eliminated. It was normalised — absorbed into how routes are priced and planned, permanently, on the assumption that it can recur.

Selective return, by cargo type

The most revealing detail in the data is that different segments made different decisions about the same water.

Tankers went back well before container ships. Product tanker transits through Suez were down only about 19 percent from 2023 levels in the last quarter of 2025, against a 45 percent decline across 2024. By April 2026, with Saudi crude loading at Yanbu and heading north, tanker transits reached 529 vessels year on year and canal revenue hit its highest level in over two years.

Container lines largely stayed away, resuming only selected services with, in the words of one analysis, extreme caution.

Same route, same threat, opposite decisions — because the economics differ. A tanker carries one cargo for one customer and the freight premium can offset the risk. A container ship carries thousands of consignments for thousands of customers on a published schedule, and a single incident disrupts an entire network rather than one voyage.

This is the same finding as Article 2, in a different substrate. The same physical constraint is severe or manageable depending entirely on what you are moving through it, and any aggregate statement about the chokepoint conceals that.

The contracts hold it in place

There is a mechanical reason recovery lags beyond anyone's judgement of the threat.

Routing is written into service contracts, insurance classifications and schedule commitments that run for a year or more. A carrier deciding to return to the canal is not making a navigational decision; it is reopening agreements with insurers, shippers and ports, several of which were renegotiated on the assumption of the longer route.

So even a genuine improvement in security arrives at a system whose paperwork points the other way, and the paperwork turns over on an annual cycle at best. Analysts expecting any return to be gradual rather than sudden are describing contract renewal cycles as much as risk appetite.

What this predicts

If risk normalises rather than clears, then the relevant forecast is not when the route reopens. It is what permanent premium the market settles on, and how long the elevated cost persists after the threat recedes.

The evidence so far suggests years rather than months, and it suggests the recovery will be gradual because a rapid one would itself be disruptive to the carriers who reorganised around the longer route.

5. What Most Analysis Gets Wrong

That open or closed is the question

It is the question that gets asked and it has almost no predictive value. Both the Red Sea and Hormuz have spent most of the last three years legally open and commercially impaired, and the gap between those two states is where every operational consequence lives. Ask instead what share of normal traffic is passing, at what insurance rate, with what transit time.

That successful rerouting means the disruption was minor

It means the opposite of what it appears to mean. The system adapted, and the adaptation costs ten to fourteen days, several hundred dollars per container, and a permanent tightening of fleet capacity. Nothing failed, so nothing was counted, and the premium became part of the baseline.

That falling attacks means normalisation is coming

Incident counts and traffic recovery have been decoupled for two years. Carriers are pricing an asymmetric risk — an expensive route against a catastrophic loss — and that asymmetry does not improve as fast as the security situation.

That a return to the canal is unambiguously good news

For shippers, it would lower costs. For carriers, releasing that much tied-up capacity at once would push rates down sharply after two years of structurally inflated sailing distances. Several parties with the ability to influence the pace of return have a commercial interest in it being slow, which is worth knowing when reading confident predictions of recovery.

6. Base, Stress and Extreme

Four paths, with our probability assessment and the condition that would falsify each. Probabilities sum to one hundred.

The fourth path is worth a sentence despite its probability. Early 2026 came close to it: the Hormuz escalation and a suspension of Suez transits overlapped, and the reroute from one ran into the disruption at the other. Chokepoints are individually substitutable and collectively much less so, which is the failure mode that reroute planning rarely models.

7. Forecast — One Year, to end-2027

Traffic stays below pre-crisis levels

Probability 0.70  ·  Confidence: Medium-High

We expect Suez Canal transits for 2027 to remain at least 15 percent below 2023 levels, even if attacks remain at their recent low frequency.

The reasoning is the asymmetry in section 4. Carriers that reorganised networks around the longer route face real costs in reorganising back, and the downside of returning early is qualitatively worse than the cost of returning late.

Second-order effect. As long as capacity stays tied up in longer voyages, freight rates stay elevated on routes with no connection to the Red Sea. Shippers between, say, Asia and South America pay part of the cost of a disruption thousands of miles away, and will attribute it to something else.

What would weaken it. Transits recovering to within 15 percent of 2023, which would indicate that the asymmetry we describe resolves faster than the last two years suggest.

8. Forecast — Three Years, to 2029

Another chokepoint event of material scale

Probability 0.60  ·  Confidence: Medium

We expect at least one further disruption at a major maritime chokepoint, lasting more than three months and affecting more than five percent of global container capacity.

This is a base-rate forecast rather than a prediction about any particular location. The last three years contained at least two such events. The candidates are well known and there is no mechanism visible that reduces the frequency.

Second-order effect. Repeated events convert rerouting from contingency into network design. A carrier that has rerouted twice stops treating the alternative as an emergency measure and starts scheduling for both routes, which costs more in normal conditions and less in disrupted ones.

What would weaken it. Three years without a sustained disruption at any major chokepoint, which would be the quietest such period in over a decade.

9. Forecast — Five Years, to 2031

Dual routing becomes standard design

Probability 0.55  ·  Confidence: Medium

We expect at least two of the largest container carriers to publicly operate dual-routing as standard network architecture rather than as contingency planning — services designed from the outset to switch between corridors without renegotiating schedules or contracts.

The logic is the one from Series I: where disruption recurs, the expensive thing stops being the disruption and becomes the reorganisation. A network built to switch pays a permanent premium and avoids a repeated one.

Second-order effect. Dual routing favours the largest carriers, who can afford duplicated port relationships and slack capacity. Smaller operators reroute reactively and absorb the reorganisation cost each time, which is the two-tier outcome this series keeps finding in different substrates.

What would weaken it. Carriers continuing to treat rerouting as an exception handled case by case, which would indicate the events are still being read as one-off.

10. Forecast — Ten Years, to 2036

The old baseline does not return

Probability 0.55  ·  Confidence: Low

We expect Suez Canal annual transits in 2035 to remain below the 2023 level.

Not because of any particular ongoing threat, but because three structural changes appear durable: route diversification is now embedded in planning, alternative capacity has been built in pipelines and port infrastructure, and trade patterns themselves have shifted in ways that reduce the volume wanting that corridor.

We hold this at low confidence and say why. A decade is long enough for a durable security settlement, for trade patterns to shift again in the opposite direction, or for canal pricing to change the calculation entirely. The mechanism is plausible and the horizon exceeds what this method can support with confidence.

What would weaken it. Transits exceeding the 2023 level at any point before 2035, which would indicate the diversion was a disruption rather than a reconfiguration.

11. Signals to Watch

—   Transit counts as a share of normal, not open-or-closed status. Weekly tracking data exists and is the only honest measure of usable capacity

—   War-risk premiums as a percentage of insured value. These moved a week before the February escalation and are the earliest available indicator

—   Divergence between cargo types through the same water. When tankers return and containers do not, that gap is the information

—   Fleet utilisation and vessel-days consumed, which is where the cost lands on routes far from the disruption

—   Pipeline and overland bypass capacity around chokepoints, which is slow to build and permanent once built

—   Whether carriers describe rerouting as contingency or as network design. The language change precedes the structural one

—   Simultaneous impairment at two chokepoints, which is the failure mode almost no reroute plan models

12. Recommendations — Individuals

This reaches most people through prices rather than through shortages, and the practical content is correspondingly modest.

Immediate — 30 days

If you are waiting on anything shipped from Asia — a vehicle, furniture, equipment, a large order — add two weeks to whatever delivery estimate you were given, and treat that as the normal case rather than a delay. Most Asia-to-Europe cargo is still travelling the long way round, and quoted transit times have not always been updated to match.

Build — 12 months

Notice that freight costs are now a permanent component of the price of imported goods rather than a temporary spike. Rates stabilised well above their pre-crisis level and have stayed there for two years. Anything you buy that crosses an ocean carries that, and it will not reverse quickly even if the security situation improves.

Position — 3 years

For anything where you have a genuine choice between an imported and a locally produced version, the calculation has moved somewhat and quietly. Freight is a larger share of delivered cost than it was in 2019, and the difference is durable rather than cyclical.

Avoid. Reading chokepoint headlines as a reason to expect empty shelves. In three years of disruption, rerouting has worked every time. The cost appears in prices and schedules, not in absence.

Why this works. Knowing that the effect arrives as a permanent premium rather than a shortage tells you what to plan for — slightly longer waits and slightly higher prices, indefinitely — which is a manageable expectation rather than an alarming one.

13. Recommendations — Business

For anyone moving physical goods, this is the substrate where preparation pays best, because alternatives genuinely exist and the constraint is speed of decision rather than availability.

Immediate — 60 days

Map your inbound and outbound flows against chokepoints. Which of your lanes pass through the Red Sea, Hormuz, Malacca, Panama or the Turkish straits, and what share of your volume does each carry? Most organisations have never drawn this, and several discover that lanes they consider unrelated share one passage.

Then establish your actual current transit times rather than your contracted ones. If your planning still uses pre-2023 Asia-to-Europe durations, your inventory assumptions are ten to fourteen days wrong, permanently.

Build — 12 months

Pre-negotiate the reroute rather than planning it. The parties who handled the last three years best were those with alternative routings already priced, carriers already contracted and customers already informed about what happens when a corridor closes. The reroute itself takes days; agreeing terms during a crisis takes weeks.

Raise inventory specifically on items that cross a chokepoint, rather than across the board. This is the opposite of the general-stockpiling advice that follows disruption headlines, and it is cheaper and more effective because it targets the actual exposure.

Position — 3 years

Treat elevated freight as structural in pricing and contracts. Two years of stable premium is not a spike waiting to reverse, and contracts written on a pre-2023 freight assumption are steadily transferring margin away from whoever wrote them.

And plan for the simultaneous case. Your reroute from one chokepoint may collide with congestion at another, which is what happened in early 2026. A contingency that assumes only one corridor is impaired at a time is the most common gap in this area.

Avoid. Treating the status of a route as the operative fact. Legally open and commercially usable are different things, and your carrier's willingness to sail is the one that determines your delivery date.

Why this works. Unlike the grid or advanced packaging, this substrate has real alternatives with substitution times measured in days. That makes preparation unusually effective here: the difference between a prepared and unprepared shipper is a week rather than a year.

14. Recommendations — Capital

The structural point is that a chokepoint disruption redistributes rather than destroys, and the direction of the redistribution is legible in advance.

Immediate — this quarter

For holdings exposed to physical trade, establish which chokepoints their flows pass through and whether their cost base assumes pre-2023 freight. A company with long-dated customer contracts priced on old freight assumptions is absorbing the premium, and that shows up in margins several quarters after the disruption that caused it.

Build — 12 months

Understand that carriers and shippers face opposite exposures. Longer voyages tie up capacity and support carrier rates; a return to the canal would release capacity and compress them. That means a security improvement is a positive for shippers and a negative for carriers, and the two are often held in the same portfolio as a single logistics exposure.

Watch the insurance layer, which prices this earliest. War-risk premiums moved a week before the February escalation, and they are published. That is a rare case of an early indicator that is both public and largely unwatched outside the industry.

Position — 3 years

The durable positions in this substrate are in alternatives rather than in the corridor: bypass pipeline capacity, port infrastructure on rerouted lanes, and the additional fleet capacity that longer voyages require. These are observations about where volume moves when a corridor is impaired, not recommendations about any instrument.

Avoid. Treating a chokepoint disruption as a shortage event. Nothing has run out in three years. The effect is a permanent operating premium and a redistribution between carriers, shippers and the holders of alternative capacity.

Why this works. Insurance pricing and transit counts are published weekly, move before events rather than after them, and are read almost exclusively by people inside the industry.

15. What Would Change Our Mind

Each forecast carries its own weakening condition. Three developments would undermine this article as a whole.

—   Traffic returns to pre-crisis levels with premiums back to pre-crisis rates. Our central claim is that risk normalises into permanent cost rather than clearing; a full return would falsify it directly.

—   A chokepoint disruption produces actual shortages rather than cost and delay. Three years of evidence say rerouting works; a genuine supply failure would mean the substitutability we describe is thinner than it appears.

—   Cargo types stop diverging. If tankers and container lines make the same decisions about the same water, the segment-specific economics we identify are not doing the work we claim.

Jurisdiction and sourcing, recorded on the third row. This article's material is genuinely global — Middle East, Europe, Asia trade lanes — which improves on the American concentration of Article 1 and the Asia-and-US concentration of Article 2. The weakness is different: maritime traffic data comes largely from commercial tracking services and industry bodies rather than official statistics, and several figures are reported ranges rather than measurements. Series II is now three rows in with three different sourcing weaknesses, which is at least an improvement on three of the same.

The Series II Table — three rows

What three rows show that two could not. The substrate that generates the most alarm has by far the best substitutability. Ships reroute in days; grid connections take years; packaging lines take longer still. Public attention tracks visibility rather than severity — a disrupted shipping lane is filmable and a connection queue is not — and the table is beginning to show how badly those two diverge.

The next row is water.

Founder's Lens

[ EDITORIAL GATE — WRITTEN BY HAND BEFORE PUBLICATION. Never generated. Replace this marker with the founder's text, or record a suspension. ]

16. Bottom Line

Two competent analyses of the same water, months apart, reported that the Strait of Hormuz was closed and that it had not closed. Neither was wrong. A chokepoint has no switch — it has a transit count, an insurance rate and a schedule, and all three move while the legal status stays put.

Three years of evidence say the same thing. The Red Sea has never been shut and traffic is still around a third below normal. Attacks fell sharply and traffic did not come back, because a carrier weighing an expensive route against losing a ship is not pricing incident frequency.

So the question that matters is not whether a route is open. It is what share of normal traffic is passing, at what insurance rate, with what transit time.

And the system adapted, every time. Ships took the Cape route within weeks. Saudi crude went west by pipeline to the Red Sea. Nothing ran out, no shelves emptied, and that success is exactly why the cost went uncounted: ten to fourteen extra days, several hundred dollars per container, war-risk premiums in the hundreds of thousands per voyage, and a fleet consumed by longer voyages that raises rates on lanes nowhere near the trouble.

A permanent operating premium, spread thinly enough that nobody adds it up.

Set against the first two rows, this substrate is the good news. Ships reroute in days. Grid connections take years. Packaging lines take longer still. The thing that produces the most alarm is the thing most easily worked around, and the things that cannot be worked around at all generate almost no coverage, because a queue is not filmable.

Which makes this the substrate where preparation actually pays. Pre-negotiate the alternative route, hold inventory on the specific items that cross a chokepoint, and use your real transit times rather than your contracted ones. The reroute takes a week. Agreeing the terms during a crisis takes a month.

Forecast record

Four forecasts, one per horizon, each with a threshold, a named source that will settle it, and a date fixed before the answer is known.

Correlation. The first and fourth share a verifier and a parent cause in Red Sea recovery, and are effectively one claim measured at two horizons. The second and third share a parent cause in disruption frequency: if no further disruption occurs, dual routing is much less likely to be adopted. So this set contains two families rather than four observations, which is the weakest independence of any set in Series II so far and is recorded rather than glossed.

Directional statements elsewhere in this article carry no threshold and are excluded from the record.

Sources

A note on a disagreement in the sources, because it is the point of the article rather than a footnote. Competent analysts described Hormuz as effectively shut and as not closed in 2026, months apart, using the same tracking data. We have not resolved that disagreement and we do not think it needs resolving: it reflects a genuine ambiguity in what closure means for a waterway that remains legally open while carrying a fraction of its former traffic. Where sources conflict on status, we report traffic share instead, because that is the quantity with operational consequences.

In this series

—   Previous: Article 2 of Series II, The Narrowest Point — computing, the second row.

—   Next: Article 4 of Series II, on water — the fourth row.

—   The method behind the Chaos Index and this series: /methodology

THRIVE IN CHAOS

Decision Intelligence for an Uncertain World

Analysis → Forecast → Recommendations  ·  Signal → Meaning → Action → Stability

Signal Over Noise  ·  thriveinchaos.ai

AI intelligence system with human editorial oversight.

Forecasts are probability-based analytical assessments, not certainties. This material supports independent judgment and does not constitute financial, legal or investment advice.