On 13 August, Hapag-Lloyd and Maersk published their second-quarter figures within hours of each other, and the industry finally got the number it had wanted since February. Hapag-Lloyd’s Liner Shipping segment swung from a first-quarter EBIT loss to a profit of USD 153 million, on group profit of USD 83 million. Maersk went further, delivering group EBITDA of USD 3.0 billion and EBIT of USD 1.6 billion.
Neither release, however, leads with the bill.
Hapag-Lloyd disclosed cost headwinds of around USD 600 million in the quarter arising from the Middle East conflict, having earlier confirmed six of its vessels trapped inside the Persian Gulf. Group EBITDA for those same three months came to USD 829 million. So the largest single drag on the business consumed three-quarters of its gross earnings, and the company still turned a profit. That is not a market recovering. That is a market being rationed.
Gemini as a natural experiment
The two partners share a network, a quarter and a freight market. Their results could hardly differ more.
Maersk’s Ocean unit lifted revenue 23% to USD 10.5 billion and drove EBIT from USD 229 million to USD 935 million, with the margin widening from 2.7% to 8.9%. Hapag-Lloyd’s Liner Shipping EBIT went the other way, falling from USD 167 million a year earlier to USD 153 million, while group profit in Hamburg dropped from USD 306 million to USD 83 million, a decline of 73%.
Same alliance, same rates, opposite direction. One caveat belongs here: Maersk’s own report attributes USD 175 million of that quarterly EBIT gain to a change in the estimated useful life of its vessels, so roughly a quarter of the improvement is an accounting revision rather than a trading one. The variable that still separates the two partners is exposure to the Gulf, not commercial skill. And Maersk names the driver itself, reporting that spot rates drove earnings while elevated Middle East costs were recovered through commercial measures. Q2 2026 container shipping was decided by geography and geopolitics, not by pricing.
The arithmetic of borrowed scarcity
The mechanism sits in plain sight. Hapag-Lloyd’s own investor report records that the Shanghai Containerized Freight Index nearly doubled between year-end 2025 and the end of June, rising to USD 3,240 per TEU from USD 1,656, and attributes the climb to higher transport costs and to Asian exporters front-loading cargo to mitigate supply-chain risk. Yet Hapag-Lloyd’s own average rate rose only 8.9% year on year, to USD 1,475 per TEU, while Maersk reported average loaded freight rates up 22% year on year on loaded volume growth of 4.1%. Two partners in one network, with reported rate increases more than twice apart, which is not what uniform pricing power looks like.
Hapag-Lloyd’s first-half average, meanwhile, came in at USD 1,406 per TEU against USD 1,411 a year earlier. The entire rate recovery sits inside three months, and those months coincide with the period when the Strait of Hormuz blockage hardened from a shock into a persistent operating condition. Hapag-Lloyd publishes the consequence as a unit-cost bridge, which is the closest thing this quarter offers to a measurement. Transport expenses per TEU including depreciation rose 6% to USD 1,432 across the half, with handling and haulage adding USD 33, bunker and emissions USD 15, and equipment and repositioning USD 13. The company traces the increase across higher bunker and emissions costs; storage and inland transport linked to the conflict; equipment repositioning also affected by it; and higher vessel charter costs alongside longer transit times and rerouting.
Maersk’s chief financial officer then says the quiet part on the record. Robert Erni told the investor call that freight rates alone added around USD 1.6 billion to Ocean EBITDA, a figure that included “compensation for higher bunker costs, elevated insurance premiums, long dwell times” alongside transshipment and network costs tied to contingency routing. Higher bunker prices, up 44% year on year, then took around USD 612 million back out. Read that again. The quarter’s largest earnings driver explicitly included cost recovery.
The quarter inside the half
The capacity mechanism is harder to measure directly than the cost side, but OOCL supplies an operational proxy, and its own comparison. Across the first half, loadable capacity rose 5.3%, liftings rose 5.2%, the load factor slipped 0.1 percentage points and average liner revenue per TEU increased just 0.2%. Within the second quarter alone, capacity grew 6.3%, liftings grew 8.8%, the load factor improved by 1.9 percentage points and revenue per TEU jumped 10.1%. Across the half, capacity and cargo moved almost in lockstep and revenue per TEU barely moved. Inside Q2, cargo growth outran capacity growth, the load factor stood 1.9 points above its year-earlier level, and revenue per TEU accelerated to a 10.1% year-on-year increase.
Our May analysis argued that volumes had returned without pricing power. Q2 2026 container shipping shows that rates can return without pricing power either.
The demand the industry has already spent
There is a competing explanation, and the carriers themselves supply it. Hapag-Lloyd’s own investor report cites Asian exporters advancing shipments to manage supply-chain risk among the drivers of the rate climb. Yang Ming says changes in tariff policy and rising energy costs drove import booking demand forward, bringing the traditional peak season with them. HMM dates its rate recovery to late May, as peak season arrived earlier than usual. ONE cites accelerated shipments ahead of tariff changes, higher fuel surcharges and inventory restocking. Four carriers therefore describe part of the second-quarter surge as a calendar effect rather than a purely contemporaneous increase in demand. That version deserves proper weight, and it is not a friendlier reading. Restocking counts as real demand, certainly. Front-loaded cargo is different: it is real current demand, but largely demand shifted from later periods rather than newly created. If shippers moved boxes in May and June that they would otherwise have moved in September, the third quarter inherits a hole.
Vincent Clerc, Maersk’s chief executive, rejects the framing outright, and his evidence is not trivial. He told the investor call that Asian demand grew 6.2% in the second quarter, that weekly volumes now sit above pre-conflict levels, and that this reflects genuine, structural growth in trade rather than cargo pulled forward from later quarters. Pressed by analysts on pre-buying ahead of tariffs, he said he saw no sign of it anywhere in the numbers. He also argues the cargo mix has shifted, with electrification components, grid equipment and data-centre cooling displacing the consumer goods that once drove Far East exports, which would make the demand profile structurally different from the one implied by the front-loading story.
Note, though, that the two explanations are not fully independent. ONE lists higher fuel surcharges among the reasons shipments were accelerated, and the carrier’s own results presentation attributes its higher operating costs directly to fuel prices driven by the Middle East conflict. The geopolitical shock worked through both sides of the market at once: it constrained supply while also encouraging some cargo to move earlier. Scarcity came partly from geopolitics; some demand came from the calendar. Neither, by itself, demonstrates durable pricing power once those conditions fade.
The year on year test
Across the half, the underlying weakness becomes clearer, and it reaches the quarter’s biggest winner. Maersk’s Ocean unit grew first-half revenue 7.0% to USD 18.7 billion, yet EBITDA fell 12% to USD 2.94 billion and EBIT fell 24% to USD 743 million. Hapag-Lloyd’s group profit swung from USD 775 million to a loss of USD 173 million, a deterioration of USD 948 million on volumes 1.5% higher, while Liner Shipping EBIT fell from USD 639 million to minus USD 21 million.
Taiwan tells the story three times over. Yang Ming produced one of the season’s sharpest reversals: second-quarter net profit reached NT$5.73 billion, with EPS of NT$1.64, nearly four times the first quarter’s NT$1.44 billion and EPS of NT$0.41. Yet first-half net profit still fell 18.2% to NT$7.17 billion, and operating profit fell 42.2% to NT$6.43 billion. Evergreen repeats the shape more starkly: first-half net profit fell 36.1% in local-currency terms to TWD 25 billion, with the operating margin narrowing to 14.8% from 22.3%, even as the second-quarter margin improved to 18.4% from 16.7%. Wan Hai looks like the exception, with first-half net profit nearly doubling to NT$19.2 billion; below the headline, operating profit fell to NT$17 billion even as local-currency revenue increased. At two of the three, net profit now exceeds operating profit. The bottom line is being supported materially from below the operating line, not by liner operations alone.
HMM held a 10.2% operating margin on first-half revenue of KRW 6.12 trillion, with net profit of KRW 765 billion. Respectable enough, until you recall that the first quarter alone saw net profit fall from KRW 740 billion to KRW 354 billion. Ocean Network Express supplies the sharpest illustration of all. The carrier lifted volumes to 3.257 million TEU and raised its average rate from USD 1,199 to USD 1,300 per TEU. More cargo, better pricing, higher revenue. Net profit nonetheless fell from USD 86 million to USD 31 million on revenue of USD 4.539 billion, a net margin of 0.7%. The USD 210 million benefit from stronger freight rates was more than offset by USD 125 million of higher bunker costs and USD 101 million of higher ship operating costs alone, even before the bridge’s other negative items are included.
One exception, and what it reveals
There is a clear outlier beyond Maersk, and it is the company that anchored the second half of our Q1 work.
CMA CGM reported second-quarter revenue of USD 15.7 billion, EBITDA of USD 3.0 billion and volumes of 6.3 million TEU, with the margin improving by 1.7 percentage points. The engine, however, was the liner business itself: shipping revenue rose 22% to USD 10.0 billion on average revenue per TEU of USD 1,575, up 15.1%, while shipping EBITDA climbed to USD 2.3 billion from USD 1.6 billion and the margin widened 3.3 points to 22.7%. Revenue from other activities, which include terminals, air cargo and media, rose 47.6% to USD 1.5 billion. Logistics revenue rose 8.5% to USD 5.0 billion, but logistics EBITDA fell 15.4% to USD 388 million.
That logistics reversal complicates the easy reading. In May we argued that terminal exposure limited the rate damage a carrier absorbed. Q2 2026 container shipping refines the finding rather than confirming it: diversification did not create CMA CGM’s rate recovery, and one diversified arm went backwards while that recovery was under way. What the adjacent businesses did was convert a shipping upturn into a broader earnings base, which makes diversification an earnings architecture rather than merely a defensive cushion. Hapag-Lloyd is moving in the same direction, having lifted Terminal & Infrastructure revenue to USD 360 million in the first half from USD 244 million, across stakes in 24 marine terminals, though that increase reflects both newly consolidated activities and volume growth. Maersk, meanwhile, keeps expanding APM Terminals, signing an agreement in the second quarter to develop the Lien Chieu terminal in Vietnam, a project representing an investment of over USD 1.7 billion and expected to handle up to 5.7 million TEU annually. Three of the largest liner groups are therefore allocating capital beyond pure liner shipping, and that pattern looks less like cyclical positioning than a structural judgement.
What Gemini is actually worth
For eighteen months the Gemini Cooperation was discussed almost entirely in operational language: schedule reliability, hub-and-spoke design, network resilience. This quarter it finally acquired a price.
Clerc confirmed that with the network fully implemented for twelve months, the Ocean cost benefit came in at about USD 950 million, just above the USD 700 million to USD 900 million range previously communicated. Hapag-Lloyd still frames the partnership in service terms, describing Gemini as delivering industry-leading schedule reliability while separately reporting a network of 129 services at the end of June. One partner describes the network; the other has now costed it.
The operational evidence puts shape around the number. Maersk says volume growth outpaced fleet growth by two percentage points on Gemini efficiencies, with utilisation running at 96%. That is the asset-turn argument stated in figures rather than theory: volumes growing faster than the fleet deployed to carry them. The company also concedes that Gemini is now fully in the base, so future uplift will be less pronounced and volume growth should converge back towards fleet growth. Gemini has reset the cost base. Unlike the quarter’s other tailwinds, that benefit is now embedded. What it cannot do is deliver the same step-change twice.
The unwinding has already started
The most consequential assumption of the season sits inside ONE’s guidance. Having raised its full-year forecast from USD 300 million to USD 900 million, ONE disclosed the conditions underpinning it: that Hormuz conditions stabilise in October, and that Cape of Good Hope rerouting continues all year.
Its Gemini competitors are not waiting. On 6 July, Maersk and Hapag-Lloyd announced a structural change to the AE15 service, moving it back to the trans-Suez route instead of the Cape of Good Hope. Maersk shares fell more than 5% in Copenhagen that morning and Hapag-Lloyd fell 3%. On 10 August, three days before results, the partners moved a second loop, AE19, back through the Red Sea.
That market reaction is the most useful number of the quarter, and no carrier published it. The first announced step back towards Suez was met by an immediate sell-off in both partners, consistent with the market pricing the risk that shorter routings would release effective capacity back into the system. Clerc’s counter-case is narrower than outright disagreement. Four services now transit the Bab el-Mandeb, about a third of the volumes that would ordinarily use the route, and he told analysts a fuller return would cut costs while having very little effect on prices, because, in his reading, the binding constraint has shifted towards congested ports and inland infrastructure. Note also that Suez and Hormuz run on separate clocks: the Red Sea is reopening in stages while disruption around Hormuz persists, so the two need not unwind together.
The case against this reading
Clerc offers the strongest counter-argument, and it deserves stating in full. He told analysts that cumulative head-haul growth out of the Far East has run at roughly 25% over three years while global terminal capacity has grown about 10%, that Shanghai now carries a twelve-day waiting time, and that a greenfield terminal takes seven to ten years from concept to operation. His conclusion is blunt: the bottleneck has moved from ships to the land side, and it cannot be cleared quickly. If he is right, congestion could keep effective capacity constrained after Hormuz reopens and after any front-loading exhausts itself, which makes it a third absorber without the explicit expiry date attached to either of the first two.
It is still not pricing power. Congestion raises costs as it raises rates, as Hapag-Lloyd’s own unit-cost bridge demonstrates, and it is not costless to the carriers either: Yang Ming reports that extended berth waiting times at Shanghai and European terminals reduced its effective fleet capacity and partially offset its revenue growth. Rates held up by congested infrastructure are rates held up by someone else’s bottleneck, which is not the same thing as commercial strength. Meanwhile the nominal supply picture has moved decisively against the carriers. Citing MDS Transmodal, Hapag-Lloyd reports that the global orderbook swelled to 12.2 million TEU by the end of June, from 10.3 million at the close of 2025, lifting the orderbook as a share of the existing fleet from 31.8% to 38.5% in six months. The industry forecasts reproduced in the same report show net capacity growth of 8.1% against container-volume growth of 3.2% for 2027, up from the 7.2% forecast reproduced in Hapag-Lloyd’s report three months earlier. Maersk’s interim report adds that the nominal fleet ended the quarter 5.4% larger than a year earlier, with demolitions close to zero for a sixth consecutive quarter.
Watch what the carriers do with the windfall, rather than what they say about it. Alongside its record second-quarter results, Wan Hai set out plans to take delivery of 42 vessels between 2027 and 2030, putting the capacity increase from its wider expansion programme at roughly 475,000 TEU from 2026 onward, including a near-USD 1 billion order placed on 12 August. Constrained effective capacity, in other words, has not stopped the industry from adding nominal capacity. Whether that capacity becomes surplus depends on which side of the argument above proves right. The supply response, however, is no longer hypothetical. The ships are being ordered now, and the commitments are already signed.
What Wednesday decides
Two sets of numbers remain outstanding. COSCO Shipping Holdings publishes first-half results at the end of the month, with OOCL already pointing the way through second-quarter liner revenue of USD 2.5 billion, up 19.8%. And on 19 August, before American markets open and once again without a conference call, ZIM reports.
ZIM is the test this analysis has been building toward. In the first quarter it was the sample’s outlier by a wide margin, posting a net loss of USD 86 million on volumes down 8%. We called that the signature of a business model that had lost its insulating capacity. The second quarter now hands that model a sharply friendlier rate environment, and the Zacks consensus estimate still points to a loss of USD 0.10 per share.
Should that estimate prove right, the commercial conclusion becomes difficult to avoid. A tide that lifts everyone except the standalone mid-tier is not a tide at all. It is a sorting mechanism.
Even the exit is contested
The escape route acquired a rival before it ran into the state. Hapag-Lloyd signed the USD 4.2 billion agreement on 16 February, and ZIM’s shareholders approved it on 30 April. Within a week, Israeli businessman Haim Sakal bid USD 4.5 billion in cash, valuing ZIM at USD 37.50 a share, a 7.1% premium to the price agreed with Hapag-Lloyd and the Israeli fund FIMI. He added a USD 250 million employee bonus pool and a pledge to keep ZIM’s 145 ships and operational headquarters under Israeli sovereignty. The shares rose 9.5%. The board replied that the merger was binding and had already won 97% support in the shareholder vote, while lawyers noted that the superior-offer clause had closed with that approval.
Two months later the objection arrived from higher up. On 6 July, ZIM shares fell 6.8% after Israeli officials came out against the sale. Prime Minister Benjamin Netanyahu reportedly told a cabinet meeting the transaction was “not on the agenda,” after Deputy Minister Almog Cohen argued that Qatari and Saudi shareholdings in the German buyer made the deal a strategic threat.
Note the date. The same Monday that Maersk announced a Gemini service returning to Suez and both partners’ shares fell, ZIM’s fell for an entirely different reason. One trading session, two geopolitical triggers, neither of them a trading result. That is the shape of this industry in 2026. The sector has not rebuilt durable pricing power this quarter; it has borrowed scarcity from geopolitics and part of its demand from the calendar. Now the same geopolitics reaches beyond earnings into ownership. The deal was engineered around precisely this sovereignty constraint: the Golden Share was intended to transfer to a new Israeli carrier created by FIMI, preserving a domestically controlled operator under the ZIM name. Shareholders approved the merger on 30 April. Completing the exit still requires the consent of the State of Israel. We return to it once those numbers are in.