Analysis

Nine Years' Notice

The Strait of Hormuz, the Red Sea, the Panama Canal, the Rhine, the Danube, the Black Sea and the Amazon are all impaired at once. The claim that nobody saw compound failure coming is false. The study was published in 2017, with recommendations attached, and the work was never done.

By xbard 4 August 2026 18 min read

In the last week of July, four notices went out that have nothing to do with each other and everything to do with each other.

On 20 July a shipping agency in Panama posted a client advisory: from 25 July the canal authority was suspending its daily auction of close-in transit slots at the Panamax locks, citing "current hydrological conditions" and the El Niño now building over the canal's watershed. On 26 July the cargo-tracking firm Kpler counted fifteen commodity vessels through the Bab el-Mandeb strait at the mouth of the Red Sea, down 56 per cent in six days, after the Houthis declared a shipping embargo against Saudi Arabia and struck four Saudi tankers to show they meant it. On 31 July Reuters reported the Rhine at Kaub, the pinch point for German industry, at roughly 25 centimetres of navigable depth, equal to the record low of 2018, with cargo vessels sailing at around a fifth of their capacity. And in mid-July Maersk issued an advisory for the Amazon river system: draft restrictions expected from October, a possible closure in November, constraints running into early 2027.

Two wars and three droughts, on four continents, in the same handful of days. Behind them stands the closure that started the year: the Strait of Hormuz, shut by Iran within days of the war beginning at the end of February, partially reopened under a ceasefire that is now visibly coming apart.

The world's maritime map has about a dozen places where trade concentrates into water a few miles wide. This summer, most of them are impaired at the same time.

There is a line doing the rounds in market commentary and in at least one widely shared video essay. We have strong historical data, it goes, for what happens when one of these chokepoints fails. We have no data and no model for several failing at once. It feels true and it is half true. The false half is the half that matters, because the scenario now unfolding was studied, modelled and published nine years ago, with recommendations attached and a warning that the work had to start immediately. The work did not start. That is not a gap in the data. It is a gap between what was known and what anyone was willing to spend, which is a political condition, not a scientific one.

The map is short

Start with the geography, because the geography is the entire setup. Oil, gas, grain and containers do not cross the ocean evenly. They funnel. The US Energy Information Administration keeps the canonical list of oil chokepoints: the Strait of Malacca is the largest, carrying 23.2 million barrels a day in the first half of 2025, with Hormuz just behind at 20.9 million. For food, the reference work counted fourteen chokepoints that matter: eight maritime passages from Panama to Malacca, three clusters of critical export ports, three inland networks of roads, rail and rivers. More than half of all internationally traded grain passes through at least one of the fourteen. For a tenth of it there is no alternative route at all.

The system's operating assumption was never that these places are safe. Everyone knows they are not. The assumption is that they fail one at a time. Every buffer in the industry is priced on that basis: the spare fleet that absorbs a rerouting, the insurance that prices a war zone, the inventories that ride out a delay. One failure means going around. Going around is expensive, slow and survivable. There is a century of data on it. The whole architecture is a single-failure design.

Malacca, the largest of them all, is so far untouched. That is the good news and it is all of it.

The ledger, without the inflation

The temptation in a week like this is to reach for the biggest number available. Some of the biggest numbers in circulation are wrong. The honest ledger is bad enough.

Hormuz. The war began on 28 February; Iran declared the strait closed within days. The ceasefire text, announced as complete on 14 June and signed on 17 June, the Islamabad Memorandum, provided for a toll-free reopening of the strait for 60 days. That window expires in the middle of August. Traffic surged in late June as the backlog cleared, then collapsed again as the ceasefire frayed. By the last weekend of July the shipbroker Clarksons counted 13 transits a day, down 90 per cent on pre-war levels, with Gulf crude exports near one million barrels a day against 15 million before the war and no gas carriers through in more than ten days. On 31 July a tanker carrying Qatari LNG was struck inside the strait, the second in a month. The figure you will hear, over 20 million barrels a day off the market, overstates it: total Hormuz flow was 20.9 million barrels a day before the war and partial flows have come and gone since June. The defensible version is the Council on Foreign Relations estimate: more than 10 million barrels a day of oil and roughly 300 million cubic metres a day of LNG, lost for over 100 days. By early July the American president was saying of his own ceasefire, "To me, I think it's over."

The Red Sea. On 20 July the Houthis declared an embargo on Saudi Arabia: all vessels loading or discharging at any Saudi port prohibited. Between 22 and 28 July they struck four Saudi tankers. Commodity-vessel transits of Bab el-Mandeb fell from 34 on the day of the declaration to 15 six days later. The strait is not closed: some 73 commodity vessels crossed over the first weekend of August, though the daily count fell from 28 to 18 across it. The embargo also visibly has a flag policy: 22 Chinese-owned, operated or flagged vessels made Saudi calls or transits in its first fortnight without being touched. What the embargo does threaten is the workaround. Saudi Arabia's East-West pipeline to Yanbu, running at its full seven million barrels a day precisely to bypass Hormuz, discharges into the Red Sea. Yanbu is still working, with five supertankers berthed there on the first of August and tankers hauling its crude out through Bab el-Mandeb with their transponders off. Plan B for the world's most famous chokepoint now empties into its most attacked one.

Suez. The canal that carries 30 per cent of global container trade never recovered from 2023-24: containership transits were still running at roughly a third of their old levels before this summer began, 1,850 last year against an average of more than 5,000 a year from 2016 to 2023. The reroute around the Cape of Good Hope adds around 4,000 miles to each journey and in the 2024 crisis it pushed Asia-Europe container rates up nearly five-fold.

Panama. A NOAA El Niño advisory is in effect; NOAA puts an 81 per cent chance on a very strong event across October to December and 23 of 26 models in the July forecast plume agree. The canal authority cut draft limits on 1 July and suspended its daily auction of close-in transit slots at the Panamax locks from 25 July. The last El Niño is the template: monthly transits fell from over a thousand to 662 in February 2024 and the canal moved 29 per cent fewer ships in fiscal 2024 than the year before.

The Rhine and the Danube. At Kaub the navigable channel fell to about 25 centimetres on 31 July, equalling the 2018 record, with the German waterways administration forecasting 24. Barges on the affected stretches are loading at around 20 per cent of capacity and the cost of tanker-barge freight from Rotterdam to Karlsruhe went from about €45 a tonne at the end of June to €150-155 five weeks later. On the Danube the Budapest gauge reached 23 centimetres, ten below the 2018 record; most cargo traffic on the Hungarian stretch has halted and Hungary's Paks nuclear plant, which generates close to half of the country's electricity, powered down entirely at the weekend for want of cooling water: the first full shutdown in its 44-year history.

The Black Sea. Russian strikes have removed about a third of Ukraine's grain export capacity, by the Ukrainian farmers' union estimate reported by Reuters in mid-July: throughput at the Odesa ports cut from roughly six million tonnes a month to four. Ukraine supplies about 6 per cent of world wheat exports and 12 per cent of corn. The damage runs both ways: loading at Sheskharis, the Novorossiysk terminal that handles roughly a fifth of Russia's seaborne crude, halted from 21 July amid Ukrainian drone attacks. The Turkish Straits carry about 3.7 million barrels of crude and products a day past both wars.

The Amazon. Brazil's Northern Arc ports moved 39 per cent of the country's corn and soybean exports in the first ten months of 2024. Maersk's advisory expects draft and capacity restrictions on the river system from October, a possible closure in November and constraints running into early 2027, on the strength of the same El Niño now drying out Panama.

Underneath all of it runs the fuel line. Very low sulphur fuel oil at Singapore touched $869 a tonne on 26 July and was still around $800 at the month's end, roughly a quarter above where it began July. That reprices every route including the healthy ones, because the Cape diversions that absorb Suez traffic burn more of the fuel that Hormuz made expensive. Drewry's world container index stood at $4,255 per 40-foot box at the end of July, just off a July peak that was its highest since September 2024, with carriers announcing emergency fuel surcharges for August. No single number here is apocalyptic. All of them are moving in the same direction at once and each one consumes a buffer the next one was counting on.

Nine years' notice

Now to the claim that nobody has data and nobody has a model.

In June 2017 Chatham House published a report called Chokepoints and Vulnerabilities in Global Food Trade. It named the fourteen chokepoints. It contained a section titled "Hazard correlation". It said, in terms: "Chokepoint disruptions do not necessarily occur in isolation... Where regional conflict or extreme climate conditions threaten local chokepoints simultaneously, the risks to food security are multiplied. For example, the greatest threat to food security among the GCC countries is some kind of regional conflagration that simultaneously disrupts two or three of the critical maritime chokepoints surrounding the Arabian peninsula." Read that sentence again, then look at the map of this summer: Hormuz, Bab el-Mandeb and the approaches to Suez, disrupted together, by a regional conflagration.

It warned about the climate half too: "Climate change may also increase the risk of concurrent supply disruptions. As extreme weather events become more common, the chances of coincidental disruptions occurring at different locations are likely to increase." It ran the compound arithmetic, modelling a US Gulf hurricane coinciding with impassable Brazilian roads and a Black Sea heatwave: around 51 per cent of global soybean shipments and 41 per cent of maize exports halted or delayed. It even named the blind spot this piece is about: "chokepoints are systematically overlooked in assessments of strategic food security."

Its recommendations were specific. Integrate chokepoint analysis into mainstream risk management and security planning. Invest in infrastructure to ensure future food security, including routes that bypass the chokepoints. Enhance confidence and predictability in global trade. Develop emergency supply-sharing arrangements and smarter strategic storage, modelled in part on the International Energy Agency's oil mechanism, with stocks pre-positioned in the most exposed regions. Build the evidence base. And on timing: "Work must begin now for the necessary measures to be in place before climate change becomes a major source of disruption and instability."

That was nine years and several governments ago. What got built? Monitoring, mostly. The IMF and Oxford launched PortWatch in November 2023, a public platform tracking daily transits at more than two dozen chokepoints and simulating disruption spillovers; the UN has adopted it. The same Oxford group published a peer-reviewed model of systemic chokepoint risk in Nature Communications in late 2025, putting the expected value of trade disrupted at chokepoints at 192 billion dollars a year. So the data claim is false and the model claim is false. What never got built was the other list: no IEA-style emergency mechanism for food, no supply-sharing arrangements, no chokepoint remit for AMIS, the G20's agricultural market monitor. By the FAO's own figures, between 20 and 30 per cent of globally traded fertilisers move through Hormuz. In May 2026, with the strait shut, the FAO warned of "a severe global food price crisis within six to 12 months" and recommended diversifying ports, corridors, storage and logistics. That is the 2017 list, reissued nine years later as urgent advice, with the window, in the FAO's own words, "closing quickly". The World Food Programme projects almost 45 million additional people falling into acute food insecurity if the conflict runs past mid-year and oil stays above $100 a barrel.

The dress rehearsal is barely two years old. In February 2024 UNCTAD's press release was headlined "Unprecedented shipping disruptions raise risk to global trade" and its standfirst named the shape: "Key shipping routes in the Red Sea, Black Sea and Panama Canal are simultaneously under threat." The rapid assessment behind it spelled out the mechanism, noting the Red Sea disruption "comes on top of disruptions that are already constraining vessel crossing in the Panama Canal and the war in Ukraine affecting activity in the Black Sea". By mid-2024 tonnage through Suez was down 70 per cent while Panama had run a third or more below peak through the spring. UNCTAD's flagship report that year was subtitled Navigating maritime chokepoints. The historical data for simultaneous failure exists. A UN agency collected it in real time, printed the word simultaneously and filed it.

The honest objection

There is a serious counter-argument and it deserves to be taken at full strength, because what it proves is worse than what it disputes.

The 2024 cascade was absorbed. ECB economists, writing that April, found "muted effects on trade and inflation": even their escalated scenario, with Red Sea disruption running to the end of 2024, added less than a fifth of a percentage point to global inflation. JP Morgan's analysts reckoned the Cape reroute soaked up about 9 per cent of effective global container capacity and expected rates to fall back quickly once the disruption ended. They were right. Rates spiked and subsided; the shelves stayed full; the great chokepoint crisis of 2024 reached most consumers as nothing more than a line in a quarterly report.

Look at why. The same UNCTAD report carries the answer in a section heading: in 2023, fleet capacity grew faster than maritime trade volumes and the longer routes helped absorb the surplus. The shock landed on a market with spare ships, weak demand and full inventories. The buffer was not designed. It was a coincidence of the shipping cycle. 2024 was absorbed because the system happened to be slack and the slack has since been eaten: the spare fleet is deployed around the Cape, the war premium is priced into every hull that goes near the Gulf, fuel has repriced and this time the failures arrive together, with an El Niño forecast to peak just as the Amazon and Panama windows tighten. Luck is not a resilience policy. It is what the absence of one looks like while it is still working.

The island at the end of every line

Now bring it home, because no economy in Europe has less business being casual about this.

The Irish Maritime Development Office's annual report, published in April, states the two governing facts in a single paragraph: Ireland's trade-to-GDP ratio is 246 per cent, among the highest in the world; approximately 90 per cent of Irish trade by volume moves by sea. Irish ports handled 57 million tonnes of cargo in 2025. Merchandise imports hit a record 49.3 million tonnes. Nearly half of all port tonnage goes through one port, Dublin, along with almost three-quarters of the State's container traffic. The National Risk Assessment itself concedes the State is "unavoidably reliant" on Dublin Port.

Ireland's exposure to this summer is not that the ships stop arriving. They have not and they are unlikely to. The IMDO itself notes that "much of the adjustment has already occurred, with most Irish container traffic being rerouted via the Cape. The main risk is a prolonged period of higher costs and tighter capacity." That is the same shape we traced in June for oil: Ireland has no buffer, so the world price walks straight through. Crude tanker earnings in the first week of the war ran at more than four and a half times their 2025 average. War-risk surcharges of $1,500 to $2,000 a box went onto containers from the East. Emergency fuel surcharges land this month. All of it arrives at the quay in Dublin marked cost, to be passed along. An Irish container supplier reported this spring that businesses here have started buying 20-foot boxes as on-site buffer storage, abandoning just-in-time by improvisation, one yard at a time. The firms are building, at retail prices, the buffer the State never specified.

The food line runs the same way. Ireland grows between 2 and 2.4 million tonnes of grain a year and imports about five million tonnes of feed materials: 1.7 million tonnes of maize last year and more than 800,000 tonnes of soybean meal, from Canada, the United States, Brazil and Argentina. Those are Atlantic routes, mostly clear of the burning chokepoints, which is exactly the point. The cargoes will arrive and they will arrive at whatever price a world short of Black Sea grain, Brazilian river capacity and cheap fuel sets for them. The ESRI has already named the channel: "The closure of the Strait of Hormuz has impacted three key inputs in the agricultural sector: fertiliser, feed and fuel." Teagasc forecasts average dairy farm income roughly halving this year. Grocery inflation has actually eased since April, 4.2 per cent in the twelve weeks to mid-July, which tells you the pass-through is a lag, not an absence. The autumn is where the lines cross: the El Niño peak, the Amazon restrictions, the Panama cuts and the expiry of the Hormuz toll-free window all sit in the same quarter.

And what has the State done? It has subsidised the symptom, competently: excise cuts built in two steps in late March and mid-April to 32 cent a litre on diesel and 27 on petrol, extended to September with a phased restoration after that. It has monitored: the IMDO briefs the Department of Transport frequently. What it has not done is name the risk. Ireland's National Risk Assessment, the document whose whole purpose is to list the strategic risks facing the State, does not contain the word shipping. Not once, in 41 pages. No chokepoint, no sea lane, no Suez, no Hormuz. That is the 2024 edition, because there is no newer edition: none has been published since September 2024 and in March the Irish Times reported that the process was under review in the Department of the Taoiseach while global tensions rose. That same month New Zealand, another small trading island at the end of long sea lines, stood up a ministerial group on economic security and supply chains, chaired by its finance minister and prompted by the attacks around Hormuz. Ireland has no equivalent body.

The enforcement arm of this indifference is the Naval Service. Eight commissioned ships, four operational, two at sea at any one time. 807 personnel against an establishment of 1,094. In 2023 the fleet delivered 520 patrol days out of a theoretical 1,600. A country that moves 90 per cent of its trade by sea patrols its own waters two hulls at a time and everything beyond them is, by long unstated arrangement, Britain's and France's problem. The State half admitted it in February, when its first National Maritime Security Strategy proposed letting French and British vessels patrol the Irish exclusive economic zone. You can defend that arrangement or attack it. What you cannot do is find it weighed in the risk register, because no published document does the arithmetic and the register where it belongs has not been updated since before the war began.

Geography is not complicated

The video essay that set off this piece closes on a line that survives all its inflated numbers: this is geography and it is not complicated. It is not. The straits are where they have always been. The rivers rise and fall with the rain. Anyone can look at the map and see that the world's food, fuel and freight pass through a dozen narrow places. Anyone who reads can find the 2017 report that said the narrow places will fail together as the climate warms and the wars return, so build the buffers now.

The world had nine years' notice and built a dashboard. Ireland had the same nine years and did not put the word shipping in its risk register. When the accounting is done for what this summer costs, at the quay, at the mart, on the forecourt, remember the order of events. The data existed. The model existed. The recommendations existed, costed, published and ignored. What did not exist was a government, almost anywhere, willing to pay for a buffer against a coincidence. Coincidence is now the weather. Resilience is a decision taken before the shock and on the evidence of its own published record, Ireland has not taken it.

An island, we are as a ship on the ocean, with no one at the helm.