COSCO SHIPPING Holdings: China's State-Backed Bet on the Global Container 中远海运控股

I. Cold Open & Roadmap

Picture the Bab-el-Mandeb Strait in the early months of 2024. It is barely eighteen miles wide at its narrowest point — a funnel of water between Yemen and Djibouti through which, in a normal year, something like a tenth of all seaborne trade squeezes on its way to or from the Suez Canal. In that winter it was not a normal year. Houthi missiles and drones were being fired at merchant ships. Warships from half a dozen navies were escorting convoys. And one by one, the world's container lines made the same calculation: turn south, run the long way around the Cape of Good Hope, add ten to fourteen days and thousands of tonnes of bunker fuel to every Asia–Europe voyage.

That decision, taken independently by every carrier for reasons of crew safety and war-risk insurance, did something no cartel could legally have done. It swallowed capacity. A fleet that sails further carries less cargo per year, and a container industry that had spent 2023 drowning in its own ships suddenly found itself short of them. Freight rates roughly tripled off their lows. And the single largest beneficiary, in absolute yuan, was a company headquartered on Shanghai's Bund whose controlling shareholder is an arm of the Chinese state.

COSCO SHIPPING Holdings 中远海运控股 (601919.SS in Shanghai, 1919.HK in Hong Kong) closed 2024 with profit attributable to shareholders of roughly $6.7 billion, more than double the prior year's, on earnings before interest and tax of about $9.54 billion.¹ The company itself was refreshingly blunt about why: an "escalating situation in the Red Sea region which led to a short supply of effective transportation capacity" had kept "market freight rate remained at a relatively high level."¹ Not a new service. Not a pricing breakthrough. A war.

Then the tide went out. For the full year 2025, revenue fell 6.14% to RMB 219.50 billion and profit attributable to shareholders dropped 37.24% to RMB 30.87 billion — about $4.32 billion — even though the company moved more boxes than ever, 27.43 million TEU, up 5.76%.²³ In the third quarter of 2025 alone, net profit fell 55.14% year on year while the China Containerised Freight Index sat 39.49% below where it had been a year earlier.⁴ By the first quarter of 2026, profit had halved again to RMB 5.88 billion on revenue of RMB 51.80 billion, with volumes still growing 6.70%.⁵⁶ Read those two facts together — volumes up, profits halved — and you have the entire investment case in miniature. COSCO sells a commodity. It does not set its own price.

This is a company of genuinely global consequence: the world's number-four container carrier by operated capacity, controlling roughly 3.6 million TEU of self-operated tonnage at the end of 2025 and around 4.84 million TEU including its orderbook.³⁶ It anchors the largest of the three vessel-sharing alliances. It owns terminals from Piraeus to Peru. It is one of very few Chinese industrial companies that is a price-setter-adjacent participant in a genuinely global market rather than a domestic champion.

It is also 45.25% controlled, directly and indirectly, by China COSCO SHIPPING Corporation Limited 中远海运集团, which is itself wholly owned by the State-owned Assets Supervision and Administration Commission 国务院国有资产监督管理委员会 SASAC.³ That ownership is not a footnote. It shapes who runs the company, how the cash gets spent, and — as of 2025 — whether a US federal agency writes rules with COSCO's name effectively on them.

So the question this piece chases is uncomfortable and specific. Is COSCO SHIPPING Holdings a world-class operator that has built durable structural advantage — scale, cost position, network density, an in-house shipyard complex, a state-sponsored cost of capital? Or is it a well-run cyclical asset, indistinguishable in its economics from four or five global peers, that happened to enjoy an extraordinary decade and is now walking into a record newbuilding wave and a US trade policy explicitly designed to make its ships more expensive to bring to America?

To answer it, we have to start where the company started — not with an entrepreneur, but with a policy.

Subscribe on LinkedIn

↑ Back to Top

II. Origins: Two State Fleets Become One

In 1961, China had almost no ocean-going merchant fleet of its own. It had a trade embargo, a Sino-Soviet split, and a need to move cargo that foreign shipowners would not always carry. China Ocean Shipping Company 中国远洋运输(集团)总公司 was founded that year as an instrument of statecraft: a national flag carrier so that Beijing's goods could move without depending on somebody else's hull. Three and a half decades later, in 1997, the state created a second one — China Shipping (Group) Company 中国海运(集团)总公司 — partly to inject competition, partly because bureaucratic empires multiply.

The nostalgia is not the point. The point is the DNA. COSCO was never a garage startup that discovered a market. It was capacity built because the state wanted capacity. Understand that and you can predict a great deal about how the modern company behaves: why it orders ships in the teeth of an oversupply cycle, why its executive directors take no pay from the listed entity, why it pays out roughly half its earnings on schedule, and why Washington treats it as a policy object rather than a commercial counterparty.

By the mid-2010s both fleets were in trouble. The container market after 2011 was a graveyard of good intentions — too many ships, too little trade growth, and a European debt crisis that had knocked the stuffing out of Asia–Europe volumes. Both Chinese groups were losing money in a business where the state had no interest in either of them failing. Beijing's answer was not restructuring in the Western sense. It was consolidation by decree.

On 4 January 2016, with State Council approval, the two groups were merged; the combined China COSCO SHIPPING Corporation Limited was formally established in Shanghai on 18 February 2016.⁷⁸ What made it extraordinary was not the headline but the plumbing. Both parents sat atop tangles of separately listed subsidiaries in Shanghai and Hong Kong, and Chinese securities law does not permit you to simply shuffle assets between listed companies. The result was one of the most complex restructurings in Chinese capital-markets history: dozens of individual asset transfers, disposals and injections executed in parallel, sorting container ships into one listed vehicle, tankers into another, terminals into a third, and finance leasing into a fourth.⁷ Container shipping and ports were consolidated into the entity now called COSCO SHIPPING Holdings; the old China COSCO and China Shipping Container Lines businesses were combined under one roof.

The arithmetic of the merger was straightforward and, at the time, modest: putting the two container lines together produced roughly a 7.7% share of global container capacity, enough to nudge past Hapag-Lloyd into fourth place worldwide.⁸ That is worth sitting with. Two national fleets, forty years of state investment, an economy that had become the world's factory — and the combination bought fourth place with a share in the single digits. Container shipping is a genuinely fragmented, genuinely global business, and no amount of domestic scale changes that.

There is a further wrinkle that Western investors routinely miss. Because both parents controlled multiple listed subsidiaries across two exchanges, the restructuring had to satisfy Shanghai and Hong Kong disclosure regimes simultaneously while moving assets between entities with different minority shareholder bases and different valuation rules. That it was executed at all, in roughly a year, says something real about the state's convening power in Chinese capital markets — a transaction of that complexity between unrelated private parties would have taken years and probably collapsed. It also says something about minority shareholder leverage in such a process, which is to say: limited.

What the merger did achieve was survival and optionality. It removed a competitor that Beijing effectively controlled anyway, cleaned up duplicated overheads and overlapping port calls, and created a balance sheet large enough to do something ambitious. It also hard-wired the ownership structure that governs the company to this day: an unlisted state parent holding a controlling but not overwhelming stake in a listed vehicle with substantial minority shareholders in both Shanghai and Hong Kong. That structure creates a permanent, unresolved tension. Minority holders want per-share returns. The controlling shareholder answers to a state whose objectives include fleet modernisation, decarbonisation, supply-chain security and employment — objectives that are not hostile to shareholder value but are not identical to it either.

The honest verdict on 2016 is that it was a defensive consolidation that worked as intended, and no more than that. Scale in a commodity industry buys you the right to stay at the table. It does not, by itself, buy you a better hand. Within eighteen months, COSCO's management would try to buy exactly that — a better hand — for $6.3 billion.

↑ Back to Top

III. The OOCL Deal: COSCO's Biggest Bet, and Whether It Overpaid

There is a particular kind of shipping company that everyone in the industry quietly admires. Not the biggest. Not the cheapest. The one whose containers show up when they said they would, whose documentation is clean, whose customer-service desk answers, and whose IT systems were built by people who understood software. In the container trades of the 2010s, that company was Orient Overseas Container Line — OOCL, the operating arm of Hong Kong-listed Orient Overseas (International) Limited 东方海外 OOIL, controlled by the family of shipping magnate C.Y. Tung and later run by his son.

Post-merger COSCO had bulk. What it did not have was that reputation, nor the depth of Transpacific network and blue-chip American shipper relationships that came with it. In July 2017, COSCO SHIPPING Holdings, alongside Shanghai International Port Group 上海国际港务(集团) SIPG, put a cash offer on the table: HK$78.67 per share, valuing OOIL at roughly $6.3 billion.⁹¹⁰ Against OOIL's last close it was a premium of around 30%; against book value it was about 1.4 times — a 40% premium to the net asset value of a company whose principal assets were, at that moment, depreciating steel in a depressed market.⁹¹⁰

The trade press argued about it immediately, and the argument is more instructive than the answer. Lloyd's List canvassed the sell side and found no consensus: Daiwa Capital Markets called the price "in line with expectations and a good deal," while Dongxing Securities described it as "higher than expected, but reasonable."⁹ The bull case rested on roughly $2.7 billion of assumed operational synergies plus cheaper funding for OOCL under a state-backed parent, and on a claim that deserves scrutiny: that "the core value of OOIL [is] not its ships, but its networks, clients, and management."⁹

That claim is either the most sophisticated thing anyone said about the deal or the most convenient. In a business where the physical product — moving a steel box from A to B — is identical across carriers, service reliability and customer relationships are the only plausible source of differentiation. If OOCL really had that, 1.4 times book was defensible. If what OOCL really had was a good decade and a good brand in a business where shippers switch carriers over $50 a box, COSCO paid a full price for goodwill that erodes.

Benchmarking the cheque

The most useful way to judge a price is against what the neighbours were paying for similar houses in the same week, and the container industry obligingly ran three large consolidations inside three years.

CMA CGM went first, bidding for Singapore's Neptune Orient Lines — parent of APL — at SGD 1.30 per share in an all-cash offer that closed on 18 July 2016 with roughly 97.83% acceptance.¹¹ That was a premium of about 49% to the unaffected share price, higher than what COSCO would later offer for OOIL, but for a very different animal: NOL was a loss-making carrier its state-linked owner wanted to exit, and the premium was the price of moving a distressed seller rather than the price of quality.

Maersk went second, agreeing to buy Germany's Hamburg Süd for EUR 3.7 billion — about $4.0 billion on an enterprise-value basis — in a deal agreed in late 2016 that finally closed on 28 November 2017, after clearing twenty-three separate regulatory jurisdictions.¹²¹³ The valuation worked out to roughly 66% of Hamburg Süd's 2016 revenues, against a basket of listed container lines then trading nearer 83% of revenue — in other words, Maersk bought below the market's own multiple, and it booked expected annual operational synergies of $350–400 million from 2019.¹²¹³

Set COSCO's deal beside those two and the shape becomes clear. COSCO wrote by far the largest cheque of the three. It paid a smaller premium to the traded share price than CMA CGM did, but it was the only one of the three buying a profitable, well-run target rather than a distressed or subscale one — which is precisely why it ended up paying above book value while Maersk paid below the sector's revenue multiple. Nor did COSCO's disclosed synergy expectation look conservative next to Maersk's: a figure in the region of $2.7 billion against Hamburg Süd's $350–400 million a year is a far more aggressive claim relative to deal size, and unlike Maersk's it was never subsequently reported against in a way outsiders could verify.⁹¹²

None of that makes the price indefensible. Quality assets cost more, and OOIL was the best-run independent carrier available at any price. But the benchmark exercise settles one thing: COSCO did not get a bargain, and it did not claim to. It paid up for a scarce asset in an industry where scarce assets have a habit of becoming ordinary the moment the cycle turns.

The deal completed on 26 July 2018, with COSCO holding 90.1% of OOIL and the combined group vaulting to third place globally, operating more than 400 vessels and over 2.9 million TEU including the orderbook.¹⁴ Critically, OOIL stayed listed in Hong Kong under stock code 316 and OOCL kept its own brand, its own commercial organisation and its own service standards — a dual-brand structure that persists to this day and that COSCO has been careful not to dissolve.¹⁵ That was a smart piece of restraint. The fastest way to destroy a premium service brand is to fold it into a volume operation.

But there was a bill attached that never appeared in the acquisition model.

OOCL owned the Long Beach Container Terminal at Middle Harbor — a modern, largely automated facility in the busiest container gateway complex in the United States. The Committee on Foreign Investment in the United States looked at a Chinese state-controlled enterprise taking ownership of critical infrastructure at a port adjacent to significant naval activity and said no.¹⁶ To get the transaction cleared, COSCO agreed to a National Security Agreement placing the terminal into a US-managed trust and committing to sell it within a defined period.¹⁶¹⁷ In April 2019 it went to an infrastructure consortium for close to $1.8 billion.¹⁷

Financially, this was survivable — arguably even fine, since the sale price was substantial and the buyer paid cash. Strategically, it was the tell. It was the first time a US national-security process reached into COSCO's balance sheet and removed an asset, and it established a precedent in Washington that COSCO-controlled infrastructure on American soil is a policy question rather than a commercial one. Seven years later, that same logic would be applied not to a single terminal but to every COSCO ship calling at every US port.

So how should an investor score the OOCL deal today? Three things are true at once. First, COSCO bought real network density and a genuinely better-regarded service brand, and it has managed the integration with unusual discipline by leaving the acquired brand alone. Second, the price was full, the synergy assumptions were the sort that get published and never audited, and COSCO has never disclosed the granular unit-cost data that would let an outsider verify whether the promised savings arrived. Third, the deal permanently tagged COSCO in Washington. The first is a real asset. The second is unproven. The third is a liability that keeps compounding — and it does so in a business whose underlying economics are far harsher than most investors appreciate.

↑ Back to Top

IV. The Core Business: How Container Shipping Actually Works, and Where COSCO Sits

Here is the simplest way to understand container shipping. Imagine an airline where every seat on every plane is identical, every airline flies the same routes on roughly the same schedule, the planes cost half a billion dollars each and take three years to build, and passengers choose purely on price. Now imagine that when demand booms, every airline simultaneously orders more planes that arrive three years later — which is to say, precisely when the boom has ended. That is the business. It is not a bad business because the operators are stupid. It is a bad business because the structure makes disciplined behaviour individually irrational.

COSCO SHIPPING Holdings is overwhelmingly this business. Container shipping and related services account for the great majority of group revenue — of the roughly RMB 233.9 billion the group turned over in 2024, container shipping supplied the vast bulk, with the terminals arm contributing a small single-digit percentage.¹⁸ Anyone modelling this company as a diversified logistics platform is modelling the wrong company.

The five forces, honestly applied

Start with rivalry, because it is the dominant force and it is brutal. Roughly half a dozen global carriers compete on the same headhaul lanes with functionally identical products. Alphaliner's mid-2026 rankings put MSC first at about 7.33 million TEU and roughly 21.6% of global capacity — a position it built by buying secondhand tonnage aggressively through the last cycle and which it now sails independently, having exited its old 2M partnership with Maersk.¹⁹ Maersk and Hapag-Lloyd operate together in the Gemini Cooperation, which held about 24% of deployed capacity on the Far East–Europe trade in February 2026.²⁰ The Ocean Alliance — CMA CGM, COSCO, Evergreen 长荣海运 and OOCL — held 32.4% on that same lane, with the Premier Alliance of ONE, HMM and Yang Ming 阳明海运 at 16.7% and non-aligned operators the remaining 5.6%.²⁰ COSCO Group with OOCL controls roughly 3.6 million TEU across some 559 vessels.¹⁹

Supplier power is meaningful but manageable: shipyards, bunker suppliers and, increasingly, charter owners can extract terms in tight markets. COSCO is unusual here because its ultimate parent group contains shipbuilding capability, and it has ordered directly from joint-venture yards — the 2023 order for twelve 24,000 TEU methanol dual-fuel vessels at roughly $2.9 billion, about $239.85 million per ship, went to Dalian COSCO KHI Ship Engineering and Nantong COSCO KHI Ship Engineering, joint ventures with Kawasaki Heavy Industries 川崎重工業.²¹ Having an in-house building channel is a genuine advantage in a hot newbuild market, when berth slots are scarce and yard prices spike. It is a much smaller advantage in a soft market, and it carries an obvious behavioural hazard: a group that owns shipyards has a structural bias toward ordering ships.

Buyer power is real and underrated. The customer base splits between large beneficial cargo owners — the Walmarts and IKEAs of the world — who negotiate annual contracts and can credibly threaten to move volume, and freight forwarders who aggregate smaller shippers and shop rates relentlessly. Neither group has meaningful switching costs. A shipper moves a box on COSCO this month and on Hapag-Lloyd next month with no retooling, no data migration, no retraining. This is why the "workflow lock-in" story that works for software companies has essentially no purchase here.

Substitutes are limited on the deep-sea lanes — air freight is an order of magnitude more expensive, and rail across Eurasia is a niche — though COSCO has been building intermodal alternatives, including a trans-oceanic land bridge service connecting North America to Central Asia across more than 20,000 kilometres door-to-door.⁶ These are interesting margin opportunities, not volume substitutes.

Barriers to entry are high in capital terms and low in competitive terms. Nobody starts a global container line from scratch. But the incumbents can and do add capacity almost without limit, which means the industry gets the costs of a capital-intensive business without the pricing protection that capital intensity is supposed to confer.

The alliance system, explained plainly

The single most important structural feature of this industry is the one least understood outside it. To offer a shipper weekly sailings from Shanghai to Rotterdam, Hamburg, Antwerp and Le Havre, a carrier needs something like a dozen very large ships dedicated to that one loop. Multiply across every trade lane and no single carrier could afford to cover the map alone. So carriers pool: Line A operates four ships on a loop, Line B operates four, Line C operates four, and each sells slots on all twelve as if they were its own.

That is a vessel-sharing agreement, and the Ocean Alliance is the largest such arrangement in the world. Founded in 2017 on a ten-year term, it was extended in February 2025 — a memorandum of understanding signed by the chief executives in Shanghai — through to the end of March 2032, covering seven major East–West trades linking Asia with North Europe, the Mediterranean, the Middle East and both US coasts, and representing about 29.1% of global TEU capacity.²²²³

Now hold that fact next to the moat question, because it cuts in an unexpected direction. Alliances mean that scale economies — the single biggest cost advantage in the industry, since a 24,000 TEU ship moves a box far cheaper than a 8,000 TEU ship — are shared industry-wide. If COSCO deploys a more efficient ship on an Ocean Alliance loop, CMA CGM and Evergreen buy slots on it. If Maersk deploys one on a Gemini loop, Hapag-Lloyd benefits. The alliance system converts what could have been proprietary scale advantage into an industry utility. In Hamilton Helmer's 7 Powers vocabulary, this is the reason "scale economies" is not the answer to the moat question for any container line. The economies exist; the exclusivity does not.

So where does COSCO's edge actually come from?

Test the candidates one at a time against evidence.

Cost position. COSCO operates a young-ish, large-ship-heavy fleet with an average age of about 13.6 years as of early 2026 and it builds in Chinese yards at Chinese prices with access to Chinese policy finance.²⁰ Its balance sheet carries a debt-to-asset ratio of 40.90% with cash of RMB 149.70 billion as of the first quarter of 2026, which is a genuinely fortress-like position for a cyclical.⁶ A low cost of capital in a capital-intensive commodity business is a real advantage — probably COSCO's most durable one. But note what kind of advantage it is: it is closer to a state-conferred cornered resource than to anything the operating business earned. It could be reduced by policy as easily as it was granted.

Network breadth and service quality. The Ocean Alliance gives COSCO a global map, and OOCL supplies a premium-service overlay. The problem for an outside analyst is verification. COSCO discloses volumes, lane mixes and freight indices, but it does not publish unit cost per TEU with the granularity that Maersk provides, and third-party schedule-reliability data has not consistently shown COSCO as the industry leader. Where evidence is thin, say so: the premium-service claim is plausible, is consistent with what shippers say about OOCL specifically, and is not independently provable from public filings.

Vertical integration. Terminals, logistics, shipbuilding channel, digital freight platforms. The digital supply-chain business generated RMB 11.53 billion of revenue in the first quarter of 2026, up 6.25% year on year — real, growing, and still modest relative to the container line.⁶ Non-maritime supply-chain revenue reached RMB 44.89 billion for full-year 2025, up 9.64%.² These are sensible diversifications into adjacent revenue with less rate beta. They are not yet large enough to change the group's earnings character.

Pricing power. On the commodity headhaul lanes, effectively none. This is the finding that matters most and it is not controversial. When the China Containerised Freight Index fell 39.49% in the third quarter of 2025, COSCO's third-quarter revenue fell 20.42% and its profit fell 55.14% — despite carrying more boxes.⁴ That is what zero pricing power looks like in an income statement: the price is exogenous, the operating leverage is enormous, and the volume growth cannot save you.

Myth versus reality

Three consensus narratives about this company deserve testing against the record.

Myth: COSCO is the shipping arm of the Chinese state, so it can undercut rivals indefinitely and take share at will. The reality is more mundane. COSCO's share of global capacity has hovered around a tenth for years while MSC — a privately held Swiss-Italian family firm with no state sponsor whatsoever — has grown to roughly double COSCO's size by buying ships aggressively in every market condition.¹⁹ If state backing conferred the power to buy share, the league table would look very different. What state backing actually confers is cheaper money and staying power, which show up in the balance sheet rather than in the market-share column.

Myth: the Ocean Alliance is a de facto cartel that stabilises pricing. Vessel-sharing agreements share ships, not commercial terms. Alliance members quote against each other daily for the same cargo, and the operating agreement — extended though it is to 2032 — governs capacity deployment, not rates.²² The strongest evidence is simply the freight index: if a bloc controlling roughly a third of Far East–Europe capacity could coordinate price, the collapse of 2025 would not have happened.⁴²⁰

Myth: growing volumes mean the business is winning. Volumes have grown every year through this cycle, including the years profit fell by a third and then by half.²⁴⁶ In a market where price is exogenous and capacity is abundant, volume growth is close to costless to achieve and tells you very little about competitive health. It is a share metric, not a profit metric, and conflating the two is the most common error investors make with this industry.

The conclusion an investor should draw is uncomfortable but clear. COSCO is a competent, well-capitalised operator with a favourable cost of capital and a strong alliance position, competing in an industry where the dominant variable is set by the collective ordering decisions of a handful of firms and by geopolitical accidents. It is well-positioned within a structurally difficult industry. That is a materially different proposition from possessing a durable competitive advantage — and it is why the smaller, steadier sibling business deserves a look before we walk the cycle.

↑ Back to Top

V. Ports: The Steadier, Smaller Sibling

Stand on the quay at Piraeus and the strategic logic writes itself. Container gantries rise over a harbour the Athenians used to fight the Persians; the ships alongside are feeding cargo into the Mediterranean and Black Sea; and the terminal operator is Chinese. COSCO first took over container operations at Piraeus in 2009 under concession, acquired 51% of the Piraeus Port Authority for €280 million in a 2016 privatisation, and raised that to 67% for a further €88 million in 2021.²⁴

Piraeus is the emblem, but the business is broader. COSCO SHIPPING Ports, separately listed in Hong Kong as 1199.HK and controlled by COSCO SHIPPING Holdings, handled 152,994,965 TEU of total throughput in 2025, up 6.2%, generating revenue of US$1,669.0 million, up 11.0%, and profit attributable to equity holders of US$312.1 million, up 1.1%.²⁵ The year before, throughput was 144.03 million TEU on revenue of about RMB 10.81 billion.¹⁸²⁵

Read those numbers carefully and the character of the business jumps out. Revenue grew 11% while attributable profit grew 1.1% and gross profit actually declined 0.3%.²⁵ Terminals are an annuity, not an escalator: concession-based, long-duration, volume-driven, with tariffs that move slowly and costs — labour, energy, concession fees — that do not. It is a steadier business than the shipping line, and a lower-return one. Total equity throughput, the figure that actually maps to economic ownership rather than to terminals COSCO merely has a stake in, was 46.85 million TEU, up 3.4%.²⁵ The gap between 153 million and 47 million is the gap between headline and substance, and investors should track the latter.

Why does this arm belong in the story at all? Two reasons, and neither is "hidden growth."

The first is vertical integration with a captive customer. A COSCO ship calling at a COSCO-affiliated terminal generates margin twice and gives the group control over berth priority and turnaround at the hubs that matter most to its own network. Terminal throughput grew 8.86% in the first quarter of 2026, outpacing the line's own 6.70% volume growth, which suggests the terminals are winning third-party business as well as feeding on group volume.⁶

The second is more delicate. Overseas terminals are also instruments of state strategy — nodes on the 一带一路 Belt and Road Initiative map, and increasingly points of geopolitical friction. Piraeus has drawn scrutiny in Brussels and Washington. Newer hubs at Chancay in Peru and Yangpu in Hainan sit at the intersection of commercial logic and national policy.⁶ For a minority shareholder, that dual purpose is a genuine analytical fact: capital may be deployed to terminals for reasons that a purely commercial operator would weigh differently, and those assets carry a political risk premium that a Rotterdam or Singapore terminal does not.

The proportionate conclusion: ports are a useful, cash-generative, slow-growing stabiliser that smooths perhaps a few points of group volatility and gives the network real operational control at key chokepoints. They do not change the fact that COSCO's share price and earnings live and die by the container freight rate. Which is exactly what the last six years have demonstrated, in both directions, with unusual violence.

↑ Back to Top

VI. The Profit Rollercoaster: Pandemic Boom, 2023 Bust, Red Sea Boom, 2025 Bust

If you wanted to design a natural experiment to prove that container shipping earnings are a function of external shocks rather than management skill, you could not do better than the period from 2020 to 2026. The management team barely changed. The strategy barely changed. The earnings moved by an order of magnitude in both directions.

The pandemic boom, 2020–2022. In the autumn of 2021, satellite photographs of the San Pedro Bay anchorage off Los Angeles and Long Beach showed something that had no modern precedent: dozens of container ships, some of them among the largest afloat, riding at anchor in orderly rows, waiting days and then weeks for a berth. Each one was a floating warehouse of goods that could not be landed, and collectively they were a supply shock disguised as a traffic jam.

Western consumers, locked down and unable to buy services, bought goods instead. Ports jammed. Ships queued for weeks off Los Angeles and Long Beach. Effective capacity collapsed not because ships disappeared but because they sat idle at anchor. Rates went vertical. In the first half of 2022 alone, COSCO's EBIT reached RMB 95.31 billion, up 92.20% year on year, with profit attributable to shareholders of RMB 64.72 billion, up 74.46% — from a half year.²⁶ For scale, that half-year profit exceeded the entire 2025 full-year result by a factor of two. The company paid out roughly half of it, declaring an interim dividend of RMB 2.01 per share against a stated 30–50% payout policy for 2022–2024.²⁶

The 2023 bust. Consumers went back to buying restaurant meals and airline tickets. Port congestion cleared, releasing all that trapped capacity at once. And the ships ordered in the euphoria of 2021 began arriving. Rates fell most of the way back to where they had started. Profit attributable to shareholders came in around $3.25 billion for the year — still profitable, which is itself notable, but roughly a fifth of the peak.¹

The Red Sea boom, 2024. Then came the rerouting described at the top of this piece. The mechanism is worth restating precisely because it is so often misdescribed as a "demand" event. Demand did not surge. Supply shrank, because the same fleet sailing a longer route delivers fewer voyages per year. Effective capacity absorption on Asia–Europe was on the order of a tenth of the global fleet. Rates tripled. In the company's own reporting currency, 2024 operating revenue reached RMB 233.86 billion, up 33.29%, EBIT reached RMB 69.95 billion, up 90.74%, and net profit reached about RMB 55.40 billion, up 95.08%.¹⁸ Lloyd's List described it, accurately, as an earnings bonanza driven by market tailwinds.²⁷

The relationship between those three growth rates is the whole lesson. Revenue grew by a third; operating profit grew by ninety percent. That gap is operating leverage in its purest form — the cost of sailing a ship from Shanghai to Rotterdam barely changes with the rate you charge for the boxes on it, so a third more revenue drops through at close to full margin. The same arithmetic runs in reverse on the way down, which is exactly what 2025 and 2026 have demonstrated.

Here is the analytical point that a shareholder letter would never make: 2024 was a year in which COSCO's profits nearly doubled because of a militia in Yemen. No product was launched. No cost programme delivered. The company executed competently in a market it did not create and could not have forecast. Treating that year's earnings as a base is a category error.

The 2025 unwind. Rerouting persisted through most of 2025 — and rates fell anyway, which is the single most important data point in this entire narrative. The first half looked resilient: revenue of RMB 109.1 billion, up 7.78%, attributable profit of RMB 17.53 billion, up 3.90%, and an interim dividend of RMB 0.56 per share.²⁸ Then the floor gave way. Third-quarter revenue fell to RMB 58.50 billion and profit to RMB 9.53 billion as the freight index collapsed by nearly two-fifths.⁴ Nine-month volumes were up 6.01% at 20.18 million TEU while nine-month profit fell 29% to RMB 27.07 billion.⁴ The company's own explanation — "changes in the supply and demand relationship in the container shipping market" — was accurate if unadorned.⁴

If the Cape of Good Hope routing was still absorbing capacity and rates still fell 40%, the conclusion is inescapable: newbuild deliveries had overwhelmed the artificial supply constraint. By September 2025, Lloyd's List reported the Shanghai–US West Coast index down to $1,460 per FEU, the lowest since July 2023, with Asia–US trades below break-even for many operators, and — damningly — evidence that the industry had stopped self-correcting: just 6,900 TEU scrapped in the first half of 2025 against 79,200 TEU in the whole of 2023, and only 0.7% of the global fleet idle versus 3.3% in 2023.²⁹

Into 2026. The first quarter delivered attributable profit of RMB 5.88 billion, down about 50% year on year, on revenue down 11%, with freight rates averaging roughly 14% below the prior-year period and the Transpacific trade suffering the sharpest deterioration.⁵⁶ Management noted that "the conflict in the Middle East has posed significant challenges to the safety and stability of global shipping," while observing that the region represents a relatively small share of global container capacity and that bookings for general cargo there had resumed.⁵ Interim results for the first half of 2026 had not been published as of 25 August 2026; COSCO has historically reported them in late August.

What should an investor take from six years of this? Three things. First, normalised earnings power is far below the headline peaks and probably somewhere in the vicinity of the trough years rather than the boom years, because the boom years each had an identifiable external cause. Second, the operating leverage runs both ways with startling force — a 40% move in the freight index produced a 55% move in quarterly profit, meaning the cost base is largely fixed and every marginal dollar of rate falls almost straight to the bottom line, or off it. Third, and most importantly: volume growth is not a defence. COSCO grew volumes in every single one of these years, including the worst ones. It did not matter.

That raises the obvious next question. If the cycle is beyond management's control, what exactly is within it — and how have the people running this company used the extraordinary cash the cycle handed them?

↑ Back to Top

VII. Management, Ownership, and Capital Allocation Under State Control

Read COSCO SHIPPING Holdings' April 2026 circular proposing the re-election of its board and you find a sentence that stops a Western investor cold. Of the chairman, it states that he "will not receive any remuneration from the Company as an executive Director," that he "did not have any interests in the Shares of the Company," and the identical language appears for the vice chairman and the general manager.³⁰

Sit with that. The three most senior executives of a company that earned roughly $4.3 billion in 2025 draw no directors' pay from the listed entity and hold no shares in it.³⁰ They are compensated and evaluated elsewhere — through the state group and the party-appointment system. Whatever else this is, it is not the Anglo-American model of aligning managers with minority shareholders through equity. It is a model in which managers are civil servants running a commercial asset.

The individuals are, to be clear, career shipping professionals rather than parachuted-in bureaucrats. 万敏 Wan Min, aged 57, chairs the board. He graduated from Shanghai Maritime College — now Shanghai Maritime University — in transportation management and engineering, took an MBA at Shanghai Jiao Tong University, and holds an engineer's qualification.³⁰³¹ His career runs through COSCO Container Lines and the parent group, but with an unusual detour: from December 2017 to October 2021 he chaired China Tourism Group, returning to become chairman and party secretary of China COSCO SHIPPING Corporation in October 2021.³⁰³² He also chairs the board of OOIL, which means the executive with ultimate responsibility for the OOCL integration also sits atop the acquired company — a governance arrangement that concentrates accountability usefully, and concentrates it in one person.

Zhang Feng, 53, is vice chairman, and his résumé is the Transpacific: general manager roles across COSCO's America trade division, president of COSCO SHIPPING (North America), vice president of COSCO SHIPPING (Southeast Asia), and today chief executive officer of OOIL alongside his COSCO deputy general manager role.³⁰ He graduated from what is now Beijing Foreign Studies University.³⁰ Tao Weidong, 55, is executive director and general manager of the listed company, chairman of COSCO SHIPPING Lines, and chairman and chief executive of Orient Overseas Container Line itself, with nearly thirty years in the group and, like Wan Min, a Shanghai Maritime degree.³⁰

What is striking, reading the three biographies together, is how little the group recruits from outside itself. All three men have spent essentially their entire working lives inside COSCO or its predecessor entities; two of the three graduated from the same maritime university; and the senior roles at the acquired Hong Kong business are now held by COSCO insiders even as the OOCL brand is preserved externally. That produces deep institutional knowledge of liner operations and near-zero fresh perspective on capital allocation — a trade-off that helps explain both the operational competence and the reflexive commitment to fleet expansion.

The pattern is coherent: a deeply experienced liner management bench with genuine international commercial exposure, running an enterprise whose ultimate objectives are set upstairs.

The capital allocation record — and it is better than the sceptics expect

Here is where the state-ownership story gets more interesting than the caricature allows. Give an ordinary management team a $16 billion windfall and the historical base rate for what happens next is grim: empire-building acquisitions, adjacent-industry diversification, aggressive buybacks near the cycle peak. COSCO did roughly none of that.

Instead the windfall went, in order of magnitude, to dividends, to deleveraging, and to fleet renewal. The dividend record is the most testable part. Against a 30–50% payout framework, the company declared RMB 2.01 per share as an interim in the boom of 2022.²⁶ In 2025 — a year in which profit fell 37% — it paid RMB 0.56 per share at the interim and proposed RMB 0.44 as a final, for total distributions of about RMB 15.41 billion, roughly 50% of attributable profit against basic earnings per share of RMB 1.99.²³²⁸

That is the credibility test, and it passed. In a down year the payout ratio held at the top of the stated range rather than being quietly cut to the bottom. The absolute dividend fell, obviously, because the earnings fell — but the policy did what a policy is supposed to do, which is behave the same way in bad weather as in good. For a state-controlled enterprise, that is a genuinely creditable outcome and it deserves to be said plainly.

The balance sheet tells the same story. Debt-to-asset at 40.90% and cash of RMB 149.70 billion against total shareholders' equity of RMB 235.69 billion is a company that used the boom to buy resilience rather than leverage.⁶ With operating cash flow of RMB 11.13 billion in a quarter when accounting profit was RMB 5.88 billion, cash generation is running well ahead of earnings — the signature of a heavily depreciating asset base.⁶

Where an activist would push

An activist investor looking at this company would not attack the dividend. They would attack four other things.

No buybacks. The shares traded through the 2025–2026 downturn at valuations that, on any normalised view, a management team with capital-allocation freedom would have found interesting. State-controlled enterprises rarely repurchase stock at scale, and the absence of that tool means the only return channel is dividends. Optionality that a Maersk or a Hapag-Lloyd possesses, COSCO effectively does not.

Procyclical ordering. The group is committing billions to newbuildings — roughly 70 dual-fuel vessels on order as of early 2026, including twelve 13,600 TEU LNG dual-fuel ships announced alongside earlier orders for twelve 18,000 TEU and six 3,000 TEU vessels, plus the 24,000 TEU methanol programme — into a market that is already oversupplied.⁶²¹ Management would argue this is fleet renewal and regulatory preparation, not gross capacity addition, and there is merit in that. But it is also the classic state-owned-enterprise pattern of investing through the trough because the mandate says modernise, and the effect on industry supply is identical regardless of the motive.

Disclosure asymmetry. COSCO reports volumes, lane mixes, revenue and profit with reasonable frequency. It does not report the unit economics — cost per TEU, bunker cost per slot, network utilisation — at the granularity that would let an investor separate operating improvement from rate movement. That gap is analytically material and it flatters the company in good years.

Complexity and related parties. A listed vehicle inside a state conglomerate, with a separately listed terminals subsidiary and a separately listed acquired carrier, transacting continuously with sister companies for chartering, agency, terminal and shipbuilding services. Every such transaction is disclosed under Hong Kong and Shanghai rules; none of them are individually alarming; collectively they mean an outside investor takes more on trust here than in a standalone carrier.

The fair summary is that management has behaved better than the structure would predict, and that the structure still limits what management can do. That limitation becomes acute the moment the constraint stops being economic and becomes political — which is exactly what happened in 2025.

↑ Back to Top

VIII. The Current Risk Radar: Tariffs, Overcapacity, and Geopolitics

On 21 April 2025, COSCO issued a statement about a US government action with a tone that shipping companies almost never use in public. "We firmly oppose the accusations and the subsequent measures," it read. "Such measures not only distort fair competition and impede the normal functioning of the global shipping industry, but also threaten its stable and sustainable development." The company warned that "ultimately, these actions risk undermining the security, resilience and orderly operation of global industrial and supply chains," and characterised the measures as discriminatory.³³

The measures in question were the outcome of a Section 301 investigation into "China's Targeting of the Maritime, Logistics, and Shipbuilding Sectors for Dominance," and they were not a generic tariff. They were a per-net-ton fee on vessels owned or operated by Chinese entities calling at US ports, with a separate and lower schedule for Chinese-built vessels operated by non-Chinese companies.³⁴³⁵ The headline rate for the Chinese-owned-and-operated category started at $50 per net ton and was scheduled to escalate annually toward $140 per net ton by 2028.³⁴

Why this was, and remains, the most consequential item on the list

Most trade-policy risks are diffuse. This one is arithmetic. HSBC modelled COSCO's annual bill at roughly $1.5 billion, equal to about 5.3% of expected 2026 revenues, with OOCL adding $654 million or roughly 7.1% of its revenues — some $2.1 billion combined.³⁶³⁷ Expressed against profit rather than revenue, the same analysis implied the fees would consume approximately 74% of COSCO's expected 2026 EBIT.³⁷ At the vessel level, a typical 50,000 net ton containership faced roughly $2.5 million per voyage in 2025, rising to about $4 million in 2026, $5.5 million in 2027 and $7 million in 2028.³⁷

There is no clean mitigation. COSCO could redeploy non-Chinese-built tonnage to US services, lean on Ocean Alliance partners to carry US-bound cargo on their hulls, or route via Canada, Mexico and Caribbean transhipment hubs — all of which were flagged as options and all of which either raise cost, cede revenue to partners, or lengthen transit.³⁶ One second-order effect deserves flagging: a strategy of preserving older non-Chinese-built vessels for US service would keep tonnage in the water that would otherwise be scrapped, worsening the industry's oversupply problem while solving one carrier's fee problem.³⁶

Management's stated response was, in effect, to eat it. In a notice to clients dated 16 September 2025, COSCO said it would absorb the additional US port fees rather than pass them to shippers, and publicly committed to "maintaining stable capacity deployment and service quality."³⁸ The line that "while the port service fees may pose certain operational challenges, COSCO SHIPPING Lines remains confident in our ability to ensure stable and reliable services in the United States" is corporate reassurance, not a mitigation plan.³⁹ Absorbing a fee is not a strategy; it is a decision about where the loss lands. It does, however, tell you something: COSCO judged that its US market share was worth more than the margin, which is a rational conclusion in a network business and a costly one.

The reprieve, and why it is not a resolution

The fees took effect on 14 October 2025.³⁴⁴⁰ They lasted twenty-seven days. Following a US-China trade and economic understanding announced on 1 November 2025, the USTR suspended the responsive actions for one year effective 12:01 a.m. Eastern time on 10 November 2025, stating that "the United States will negotiate with China pursuant to Section 301 regarding the issues raised in this investigation."⁴⁰ No party accrues liability for the fees under the relevant annexes during the suspension, and China simultaneously suspended its retaliatory port charges on US vessels.³⁵⁴¹ Notably, restrictions on LNG transport services under a separate annex were not permanently suspended and remain scheduled to take effect from April 2028.³⁵

As this is written in August 2026, that suspension has roughly three months left to run. The USTR has said it will consider before the deadline whether to extend the suspension or take further action.³⁵ For an investor, this is the defining binary of the next twelve months: a policy capable of consuming most of a year's operating profit sits dormant, its fate determined by a bilateral negotiation over which the company has no influence whatsoever. That is not a risk that can be hedged, diversified or managed. It can only be sized.

Structural overcapacity: the risk COSCO helped build

The second risk is one the industry inflicted on itself. By March 2026, the containership orderbook had reached a record of roughly 11.8 million TEU across more than 1,350 ships, equivalent to about 34% of existing fleet capacity, with Clarksons Securities placing the orderbook-to-fleet ratio at 31.6% against 27.5% in 2023.²⁹⁴² Global containership capacity had already expanded 19% since the third quarter of 2023.²⁹ Deliveries were forecast to step down through 2026 before surging again — around 2.8 million TEU in 2027 and 3.5 million TEU in 2028.²⁹

The mechanism connecting this to COSCO's P&L is direct and needs no interpretation. Supply grows faster than demand; the marginal ship must be filled; the marginal box is priced at whatever clears; the freight index falls; and because the cost base is fixed, the fall lands almost entirely on operating profit. COSCO's own 2026 framing acknowledged the shape of it: demand growth of roughly 2.5% against supply growth of about 3.8%.⁴³ That 3.8% is described as the lowest in three years, which is management's way of saying the worst is behind us on the supply side — a claim the 2027–2028 delivery schedule directly contradicts.

And COSCO is a contributor, not a bystander. Every dual-fuel newbuilding it takes delivery of adds slots to a market with too many. The prisoner's dilemma is exact: if COSCO alone restrained ordering, it would lose share to peers who did not, and the industry's rates would barely improve. So nobody restrains. This is why capacity-management levers — blank sailings, idling, slow steaming, scrapping — get deployed in downturns, and why the 2025 scrapping and idling data cited earlier is so discouraging. Carriers entered this downturn with the largest liquidity buffers in their history, and liquidity is precisely what allows an industry to postpone the discipline it needs.

Red Sea reversal, and the perverse risk of peace

The third mechanism is the one that would strike most people as backwards: a return to normal is bad for earnings. If the Suez route reopens fully, the fleet sails shorter voyages, effective capacity expands by something close to a tenth overnight, and rates fall further into an already oversupplied market.

Through 2026 this has been happening gradually and unevenly. Maersk completed a first transit in late 2025 and resumed Red Sea and Suez passages on one service by mid-February 2026, while CMA CGM kept a number of services on the longer Cape route on grounds of schedule predictability.⁴⁴ War-risk insurance premiums have remained elevated and analysts warned that a broader reopening would "aggravate an existing structural oversupply of vessel capacity."⁴⁴ COSCO's own first-quarter commentary about Middle East conflict challenging the safety and stability of global shipping suggests continued caution.⁵ The investor's takeaway is deliberately counterintuitive: the single most bullish geopolitical development for world trade would be, in the near term, bearish for this stock's earnings.

A brief second-layer note that rarely makes the headlines: the fee structure is levied per net ton of vessel, not per container carried. That design detail matters enormously, because it penalises exactly the large, efficient ships that carry COSCO's lowest unit costs and rewards the deployment of smaller, older tonnage on US services — a perverse outcome for a policy nominally concerned with maritime capability, and a direct tax on the one scale advantage the company genuinely possesses.³⁴

Green transition execution

Finally, the fuel bet. COSCO has committed heavily to methanol dual-fuel and LNG dual-fuel tonnage — 42 green ships totalling nearly 780,000 TEU in operation or under construction at the end of 2025, growing to roughly 70 dual-fuel vessels on order by early 2026, including the $2.9 billion methanol programme.³⁶²¹ It completed Hong Kong's first green methanol bunkering operation.⁶

The strategic logic is sound: IMO carbon rules are tightening, European emissions trading already prices marine carbon, and large customers increasingly demand low-carbon options. But the execution risk is real and specific. Green methanol supply at scale does not yet exist at the prices these vessel economics assume; the bunkering infrastructure is embryonic; and if the regulatory path or the fuel-price path diverges from the assumption, COSCO is left with premium-priced ships burning conventional fuel and carrying the extra capital cost. Dual-fuel is a hedge, not a bet, which limits the downside — but it is a hedge purchased at roughly $240 million per ship.²¹

Put together, these four mechanisms explain why the market has treated COSCO's post-2024 earnings with scepticism. Each maps to a specific line: Section 301 to US-lane revenue and operating profit; overcapacity to the freight rate; Red Sea normalisation to effective supply; green capex to the balance sheet and fuel cost. None of them is a generic macro worry, and none of them is within management's control.

↑ Back to Top

IX. Playbook: What COSCO's Story Teaches About State-Backed Industrials and Cyclical Businesses

Every good business story leaves behind something portable. COSCO's leaves three lessons that generalise well beyond container shipping.

Scale bought through M&A in a commodity industry purchases breadth, not power. The OOCL acquisition was strategically coherent and reasonably well executed. It bought Transpacific network density, a genuinely respected service brand, and a management culture worth preserving. What it did not buy — could not have bought — was the ability to charge more per box than the next carrier when rates are falling. The proof is empirical and brutal: COSCO owned OOCL through the 2023 collapse, through the 2025 collapse, and through the halving of first-quarter 2026 profit. In industries where the product is physically identical and switching costs are near zero, consolidation improves your cost base and your map. It does not create pricing power unless it approaches monopoly, and container shipping — with a leader at roughly a fifth of global capacity and three alliance blocs competing on every lane — is nowhere close.

State ownership is genuinely double-edged, and both edges are sharp. The favourable edge is real and it should not be dismissed by investors who reflexively discount state-owned enterprises. State backing delivered a low cost of capital in a capital-hungry industry, a shipbuilding channel most peers lack, and — most surprisingly — a disciplined, dividend-first response to a historic windfall, with the payout ratio holding near the top of its stated range even as profits fell by more than a third. That is better behaviour than many listed Western industrials managed with the same opportunity.

The unfavourable edge is that the company's asset footprint and addressable market are hostage to a bilateral relationship it cannot influence. Long Beach was taken off the balance sheet by a national-security process. Section 301 fees were designed, in substance, around companies with COSCO's exact ownership profile, and their suspension was not won by COSCO's lobbying or mitigation planning — it arrived as a by-product of a leaders' meeting. A purely commercial carrier operating identical ships on identical routes does not carry this exposure. The lesson generalises: when the state is your controlling shareholder, the state is also your risk factor, and the two are not separable.

In businesses with steep operating leverage and multi-year capacity lead times, today's record profit is tomorrow's overcapacity — mechanically. This is the deepest lesson and the one investors in every capital-intensive cyclical should carry: semiconductors, shipping, mining, refining, memory, offshore drilling. The chain is deterministic. Extraordinary profits generate extraordinary cash. Cash plus optimism generates orders. Orders take two to three years to become steel. The steel arrives into a market that has, by then, normalised. The industry's own success is the direct mechanical cause of its next downturn. COSCO's record 2024 helped fund the newbuilding wave now pressing on 2026 and 2027 rates, and every one of its competitors did the same thing at the same time for the same reasons.

The practical corollary for an investor is that in these industries, the peak-earnings year is the worst possible base for extrapolation, and the trough is the best time to ask whether a company has the balance sheet to reach the other side. On that second question — the balance sheet question — COSCO scores well. On the first, the discipline required is simply to refuse to anchor on 2024.

Which brings us to the argument as it actually stands today.

↑ Back to Top

X. Bull vs. Bear: The Investment Case Today

The bull case

Start with the balance sheet, because in a cyclical it is the precondition for everything else. Cash of nearly RMB 150 billion against a 40.90% debt-to-asset ratio and shareholders' equity of RMB 235.69 billion means COSCO can lose money for a long time without existential stress, buy assets when weaker competitors are forced sellers, and keep paying dividends through a downturn.⁶ Many of its smaller competitors cannot say all three.

Second, the payout discipline has been demonstrated rather than promised. A company that maintained roughly a 50% payout ratio into a 37% profit decline has established a behavioural track record, and that record is the closest thing an investor has to a floor under the return profile.²³

Third, the volume franchise is intact and growing. Container volumes rose 5.76% in 2025 and 6.70% in the first quarter of 2026, with terminal throughput up 8.86%.²⁶ Lane growth in 2025 was led by Asia–Europe, up 6.07%, and "other international routes," up 7.83% — the emerging-market and South–South trades that are structurally faster-growing than the mature Transpacific.² Full-year 2025 lane volumes ran roughly 4.47 million TEU on Asia–Europe, 4.16 million on Transpacific and 2.91 million intra-Asia, a genuinely diversified mix that reduces dependence on the single lane most exposed to US policy.³ Chinese export composition is also shifting toward higher-value manufactured goods, which travel in boxes and are less price-elastic than commodity freight.

Fourth, the Ocean Alliance position is locked in through March 2032, providing network breadth without the capital cost of owning every ship on every loop — and providing, incidentally, a ready-made mechanism for shifting US-bound cargo onto partner tonnage should the port fees return.²²²³

Fifth, the green fleet is a hedge against regulatory tightening that many peers have been slower to buy, and the in-house building channel means COSCO can renew tonnage on terms unavailable to carriers bidding for scarce yard slots.³²¹

The bear case

The bear case is not that any of the above is false. It is that none of it addresses the two things that determine the outcome.

The first is the freight rate, and the supply picture governing it is the worst in a decade: a record orderbook of roughly 11.8 million TEU at about a third of the existing fleet, deliveries re-accelerating in 2027 and 2028, and an industry that has demonstrably stopped scrapping and idling because its balance sheets are too strong to force discipline.²⁹⁴² A reopened Suez would make it worse. COSCO's own supply-demand framing for 2026 — demand around 2.5%, supply around 3.8% — concedes the direction.⁴³

The second is Section 301, and the honest description is that it is a suspended sentence rather than an acquittal. A fee structure capable of consuming a majority of a year's operating profit, escalating annually toward 2028, targeted at COSCO's specific ownership characteristics, currently paused until November 2026 pending negotiations to which the company is not a party.³⁵³⁷⁴⁰ Management's declared response — absorb the cost — protects volume at the direct expense of margin.³⁸

Layer on the governance constraints. No buyback capability at scale. Executive directors with no equity in the listed company and no directors' remuneration from it, compensated through a state system whose objective function includes fleet modernisation and national policy alongside returns.³⁰ Continued newbuilding into oversupply. Disclosure that does not permit outside verification of unit-cost improvement.

And then the hardest point, the one that should anchor any bear thesis: earnings quality is genuinely poor. The 2024 doubling came from a shipping-lane disruption. The 2025 decline came from rate normalisation. The 2026 halving came from rates again. Across the entire period, the variable that moved earnings was one the company does not influence. When you strip out the rate cycle, what is left is a competent operator growing volumes at mid-single digits in a market growing at low single digits — respectable share gains, no evident pricing advantage.

Applying the frameworks

Through Porter's lens, four of the five forces are unfavourable and one is neutral. Rivalry is intense and structurally so. Buyer power is high with negligible switching costs. Supplier power is moderate, partially neutralised by vertical integration. Entry barriers are high in capital but ineffective as protection, since incumbents supply the destructive capacity themselves. Substitutes are the only benign force.

Through Helmer's 7 Powers, the test is more revealing. Scale economies exist but are shared through the alliance system, which is the crucial disqualification. Network economies do not apply — a shipper gains nothing from other shippers using COSCO. Counter-positioning does not apply; there is no business model peers cannot copy. Switching costs are close to zero on commodity lanes, with modest stickiness where OOCL has embedded itself in a shipper's operations. Branding has some substance in the OOCL name, worth perhaps a small rate premium, not a durable moat. Process power is the most plausible candidate — accumulated operational capability in network design, terminal integration and increasingly in digital freight — but COSCO's disclosure does not allow it to be verified against peers. Cornered resource is where the real answer lies, and it is not a commercial one: privileged access to state capital, domestic shipbuilding capacity, and the world's largest export origin market. That is a genuine power. It is also the same characteristic that makes the company a target of US policy. The moat and the vulnerability are the same fact.

The stress test an activist would run

A sceptical investor would ask five questions and would find that COSCO answers three of them well. Is the dividend policy real? Yes, demonstrated through a down year. Is the balance sheet sound? Yes, conspicuously. Is the acquired business being managed sensibly? Apparently yes, on the evidence of brand preservation and OOIL's continued separate listing.

The two it answers badly: Why is the company adding capacity into a record orderbook, and what return does management expect on those hulls given the rate environment they themselves forecast? And why does a company earning billions disclose so little about the unit economics that would let shareholders distinguish operating improvement from market luck?

What to actually track

Three metrics carry nearly all the information about this business, and none of them requires a spreadsheet model.

One: the container freight rate indices — the SCFI and CCFI published by the Shanghai Shipping Exchange 上海航运交易所. These are the price of COSCO's product and they are published weekly. Because the cost base is largely fixed, the direction and magnitude of these indices predict earnings direction with a fidelity no other input matches. The 39.49% CCFI decline in the third quarter of 2025 and the 55.14% profit decline that accompanied it are the calibration.⁴

Two: the industry orderbook-to-fleet ratio, and the pair of behavioural indicators that sit alongside it — scrapping volume and the idle fleet percentage. The orderbook tells you how much supply is coming; scrapping and idling tell you whether the industry is willing to discipline itself. A ratio falling from the low-thirties toward the high teens, accompanied by rising scrapping, would be the single clearest signal that the cycle is turning. Its absence is the clearest signal that it is not.

Three: the status of the Section 301 action after the suspension expires in November 2026. Reinstatement, extension, or permanent withdrawal — this is a binary policy event with an estimated impact equal to a majority of a year's operating profit, and it will be decided by governments, not by the company.³⁵³⁷

Volume growth, throughput, and even the dividend are second-order next to these three. A reader who tracks only these will understand this company's results before the company reports them.

The open question

So which is it? COSCO SHIPPING Holdings has a fortress balance sheet, a demonstrated dividend discipline unusual for a state-controlled enterprise, the largest alliance position in global container shipping, a genuine cost-of-capital advantage, and a management team of career liner professionals who have not done anything foolish with a historic windfall. It also sells an undifferentiated product into a market with a record orderbook, has no demonstrable pricing power, discloses too little to prove the operational edge it implies, is adding capacity into oversupply, and faces a US trade action explicitly constructed around its ownership that would consume most of a year's profit if it returns in November.

The bull and the bear are not really arguing about facts here. They are arguing about which fact dominates. If container shipping is a cyclical in which the well-capitalised, low-cost, politically-protected operator compounds through the trough and takes share from weaker hands, this is a business that survives whatever 2027 and 2028 deliver. If instead the industry's own newbuilding has broken the cycle's normal self-correction — and if Washington decides in November that the reprieve has run its course — then a company whose profits are the residual between an exogenous price and a fixed cost base has very little left to defend itself with.

Nothing in the last six years of results tells you which. That, more than any number in this piece, is the honest state of the argument.