China's State Council Approves 15th Five-Year Plan for Port Modernisation Covering 24 Ports

The Analysis

Put the map of China's coastline on the table and look at it the way a port planner does, not the way a booking desk does. On 31 July 2026 the State Council Information Office confirmed it: a Fifteenth Five-Year Plan for Port Modernisation, approved by the State Council and jointly issued by the General Administration of Customs and the National Development and Reform Commission, with 24 ports written into its construction project pipeline. Twenty-four was the number everyone repeated that afternoon. It is not the number that will decide your next contract.

What will decide it is the direction behind the list, and two figures released in the same breath say more than the headline. In the first half of 2026, investment in China's water transport sector rose 19.8 per cent year on year, and investment in air transport rose 11.0 per cent. Sea nodes and air nodes, the two ends of a single network, being poured with money at the same time. Read that as a statement about the shape of the next decade rather than the shape of this quarter.

I spent years on the planning side before I moved to the trade side, so I have a habit of reading plans for what they leave unsaid. A modernisation programme framed around expanding opening-up, deepening international cooperation and serving Chinese modernisation is a drawing of a network. A drawing becomes a machine only when a great many execution premises hold, and where exactly the binding edges sit is a question for Dr. Ingrid Voss to answer properly. What I can do here is tell you which nodes are about to change shape, and how to position cargo while the map is still being redrawn.

Put the map from 2016 next to the map from 2026 and the first thing that jumps is not the number of cranes. It is the direction the arrows point. A decade ago the Chinese port conversation was still largely about water and steel: how deep the channel was, how long the quay ran, how many of the newest and largest ships could be parked alongside at once, how fast a terminal could turn a vessel around and hand the boxes to a truck. The frontier then was stacking more capacity on the same coastline, and the coastline was where the story stopped.

The frontier now has moved inland and upward at the same time. Inland, because the box that lands in Shenzhen or Ningbo now has to reach a factory park in Hunan or a distribution centre in Sichuan through a rail or barge leg that is part of the promise rather than an afterthought; upward, because a growing share of the value crossing the border travels on an aircraft, in parcels and components that will never see the deck of a container ship. A plan that writes 24 ports into its construction pipeline is not a plan about 24 pieces of concrete.

It is a plan about which nodes in a continental network will be the strong ones in 2030, and which will quietly become the second choice. Twenty-four is a short list out of the several hundred places on this map where a vessel, a train, a truck or an aircraft can cross a border. Short lists are where the money goes and where the politics of a network is settled.

Now translate the two investment numbers, because that is the part of the story a shipper can act on. Water transport investment up 19.8 per cent in the first half of 2026, air transport up 11.0 per cent. Put them side by side and a shape appears. The larger percentage sits with water, which is where the tonnage still is, and water investment is rarely just a bigger berth. It is the barge channel, the automated yard, the rail spur, the road flyover that lets a gate actually process the boxes it promises to process.

The smaller percentage sits with air, and air is the node most exposed to cross-border e-commerce, to high-value electronics, to the parcels that no longer wait for a ship and to the components whose value per kilogram makes a seven-day sea leg unthinkable. The two figures together say the planners are not choosing between sea and air. They are building the seam between them, the transfer points where cargo can move from one mode to the other without losing three days in the handover.

For an exporter, the seam is where time is lost and found, and a plan that funds both ends is a plan that will, in time, make the seam faster. In time is the operative phrase. Investment is a leading indicator, and it becomes throughput on the schedule its own contractors keep, not on yours.

There is a second reading of the plan that a shipper should hold lightly. When a document is framed around expanding opening-up, deepening international cooperation and serving Chinese modernisation, it is speaking to an audience wider than the freight market. It is a statement about which corridors China intends to keep open and competitive as the pattern of world trade rearranges itself around it: the overland links toward Central Asia and Europe, the maritime express routes toward Southeast Asia and the Gulf, the air lanes that carry a rising share of cross-border e-commerce into Europe and North America.

For a company that has been quietly diversifying sourcing across Vietnam, Thailand, Mexico and Turkey, that framing matters, because a port network is useful to you only at the nodes where your cargo actually starts or finishes. A modernised node on a corridor you do not use is a museum. A modest node on the corridor you do use is an opportunity. The plan tells you where the state intends to concentrate, and concentration is a wind that will either push your lanes forward or leave them exactly where they are.

Here I have to slow down, because I have watched too many good plans be read as finished machines, and that misreading costs real money. A modernisation programme states an intent. It does not state the day a new yard opens, the sailing a new berth will serve, the rule a joint-venture terminal will apply to a late gate-in, or the clearance time a rebuilt border crossing will achieve on its worst Tuesday. Whether one of the 24 listed ports actually gains a new direct call, whether an inland link shortens a door-to-door time, whether a corridor reallocates capacity toward exports and away from transhipment, are all execution questions.

Where the plan binds and where it merely gestures is a boundary better drawn by Dr. Ingrid Voss, who follows the rule-making side far more closely than I do. My advice is to treat the 24 names as a portfolio of bets rather than a schedule of deliveries. That single shift in reading turns the document from cheerleading into something you can route cargo against.

History helps here, if you hold it loosely. The nodes that reshaped trade in the last century did so not by being busy but by changing where the seam between modes sat. A canal that cut a continent in half, a rail link that joined a landlocked factory belt to a deepwater quay, a free zone that let goods be reworked before they were re-exported: each one moved the point at which cargo changed hands, and each one left some older node suddenly peripheral. When I look at a list like this one, I am not looking for the biggest names, because the biggest names were strong already and will stay strong whatever a plan says.

I am looking for the mid-sized node sitting on a corridor where something else is changing at the same time: a new rail line, a new cross-border economic zone, a new direct service. A node that gets both a construction budget and a new corridor is a node whose name will be on a routing slide three years from now. A node that gets only concrete just gets slightly faster at what it already did.

Before the arithmetic, be precise about what a node actually is, because the word hides three different machines. A node has a water layer, the berth and the channel; a land layer, the yard, the gate and the road or rail behind it; and a rule layer, the working hours, the cut-off discipline, the way a late box is treated. Modernisation money can touch any one of the three and they are not interchangeable. A deeper channel helps the largest ships and does nothing for a truck queuing at a gate.

An automated yard speeds up stacking and can still bottleneck at the gate if the road outside was not widened in the same programme. The rule layer, the quietest of the three, can change the effective capacity of a node more than any crane, because a terminal that stops accepting boxes outside a window is not smaller, it is differently shaped. When you read a list like this, ask which layer the money is aimed at and match it to the layer where your own pain sits.

If your problem is missed cut-offs at the gate, a berth that takes a bigger ship solves nothing for you. If your problem is transhipment timing, a widened road out front solves nothing either. The document will not tell you which layer it is funding at each of the 24 names, but your last six months of exception reports will, if you read them next to it.

Let me put a number on that portfolio so it stops being a slogan. The assumptions here are mine and I will state them plainly, because not one of them comes from the plan. Assume a mid-sized exporter moves 200 ocean containers a month, and assume 60 per cent of them, 120 boxes, run on lanes whose origin or destination node appears on the 24-port list. Assume, further, that a completed modernisation removes one day of average schedule variability from each leg that touches such a node. That second assumption covers the combined effect of a better berth, a better yard and a better inland link; it is a reasonable order of magnitude and nothing more.

Multiply: 120 boxes times one day is 120 container-days of variability taken out of your book every month, which is 1,440 container-days across the year. Now put that back into inventory. If you presently hold ten days of cover against that variability on the affected lanes, a one-day improvement lets you hold nine, and one day on 120 boxes a month is 120 container-days of working capital and yard space you no longer pay to park. Stretch the same logic over a three-year build and the released number compounds, which is exactly why it belongs in your three-year sourcing map and not in your next quotation.

There is a parallel arithmetic on the air side. If air freight is 15 per cent of your total volume today, and the air investment figure is running at a little over half the water figure, then the two modes are being grown into one system at the planning level. A company still buying air and sea from two teams with two scorecards will lag the planners by the width of that seam. Run the air number on its own once, so the two decisions sit on one page: 15 per cent of a 1,000-tonne annual book is 150 tonnes a year, and anything that shifts a tenth of that share from air to a sea-air combination changes both your spend and your promise. The plan does not set the split for you. It only builds both ends of the seam that makes the split worth reconsidering.

The plan was announced on 31 July, so make the week beginning 3 August the week you act on it. Give it an owner first: the network-design lead, or whoever owns three-year routing, not the coordinator who books this week's sailings, because this is a ten-year question triggered by a three-day news cycle. By 7 August, that owner should produce one sheet listing every lane you run, the node at each end, and a note on whether that node is on the 24-port list, on a corridor adjacent to one, or on neither. Score each lane twice, once on the share of your volume that clears through a listed node and once on the transit variability you currently carry.

Then write a threshold and hold to it: any lane where more than half your volume runs through a listed node and where on-time performance sits below 92 per cent becomes a candidate for a second routing option before the year ends. Do not attempt all of them. Choose three lanes and build the alternative properly. By 31 August, ask your forwarder for direct-call evidence on those three lanes, meaning the actual vessel rotation that touches the modernised node and not a brochure line about potential. Run one live test booking on the best of them during September, so the alternative is proven rather than assumed; an untested backup node is just another way of saying you have one route.

In the same window, read your routing clause. If it fixes only an origin port and a destination port, add a named alternative node per lane and the trigger that switches you to it, so the decision is written before you need it and is not renegotiated the morning a gate closes. Finally, once a quarter, put the two figures back on the table, water at 19.8 per cent and air at 11.0 per cent this half, and ask the only question that matters across a decade: does the way you split cargo between sea and air still match the way the network is being built?

Three choices are worth pricing against each other, and each carries a cost the plan will not pay for you. Move cargo onto a newly strengthened node and you buy resilience, but you inherit that node's learning curve: the first sailings through a modernised terminal are not always the fastest, and a rail or barge leg that is new to your company adds its own documents and its own people to train. Stay on the established node and simply buy more buffer inventory, and you trade working capital for certainty, which is a perfectly legitimate answer on lanes whose product can carry the cost. Split the lane between a primary and a secondary node and you pay in complexity: two booking windows, two sets of gate rules, two document streams, and a routing call that has to be made earlier than you are used to making it.

The pitfalls gather where they always gather. The first is reading a plan as capacity that already exists, treating 24 names as berths available this quarter rather than a construction pipeline measured in years, and then building a customer promise on a promise someone else made to a contractor. The second is documents that stop following the cargo the moment it moves: origin declarations, certificates and the real routing diverge on a secondary node, the goods are held for the paperwork, and the savings are eaten by storage and re-handling. The third is organisational, and it is the quiet killer, leaving the network decision with the team that books daily, because a team measured on this week will always choose this week's node, and the whole point of the 19.8 per cent and the 11.0 per cent is that they describe a horizon that team cannot see. The fourth is subtler still, and it is the one I most want to warn against. Do not assume the plan does your thinking for you. It does not remove the need to decide. It changes the field on which you decide.

Ports are among the slowest things people build. A quay takes years, a channel takes a decade, and the shape of a trade route can outlive the government that dredged it. That is why a five-year plan for modernisation is less a piece of news than a piece of weather. It tells you which way the current is running, and you set your sails to it whether or not you like the forecast. The 24 ports on that list will look obvious in 2035 and strange in 2026, the way the busiest nodes of the last decade looked odd to everyone who first heard their names.

If you want one thing to watch over the next few years, watch the seam between sea and air. That is where the network is being rebuilt, and that, more than any single berth, is where your next decade of routing is quietly being decided. The list will be revised, some names will drop out and others will appear, and the two investment figures will rise and fall with the construction cycle. What will not change is the direction: a network that is deliberately knitting its sea end and its air end into one thing, and asking every shipper to decide, sooner or later, where on that seam they want to stand. Decide early and you are choosing. Decide late and the choice is made for you by whoever gated in first.

↑ Back to the key points

— By Tomás Ferreira