MSC Books 20 More 24,000 TEU Ships at Hengli as Orderbook Hits 46.5% of Fleet

The Analysis

Linerlytica's note says MSC has booked 20 more 24,000 teu ships at Hengli Heavy Industries in China. That is 480,000 teu in a single order, and it takes MSC's programme of ships above 20,000 teu at that one yard to 62 hulls. The same note puts the global containership orderbook at 16.2m teu, equal to 46.5% of the existing fleet, the highest that ratio has been since 2009. MSC operates about 7.4m teu, 21.5% of world liner capacity, with 2.94m teu on order.

We check the arithmetic before arguing about the meaning, because two of these numbers have to reconcile and they do. If 16.2m teu is 46.5% of the fleet, the fleet is about 34.8m teu. MSC's 7.4m teu against 34.8m is 21.3%, close enough to the 21.5% quoted that both figures are describing the same fleet. The set is internally consistent, which is more than you can usually say about orderbook reporting.

Now the part the headline does not test. MSC's 2.94m teu on order is 39.7% of its own 7.4m operating fleet. The industry's 16.2m teu is 46.5% of its fleet. So the carrier everyone points at as the driver of a record orderbook is ordering less aggressively than the fleet average, measured the only way that is fair, which is against your own size. MSC is 18.1% of the global orderbook against 21.5% of the fleet, so it is under-indexed on newbuilds. That does not make MSC cautious. It means the record is being set by everybody else, and that MSC's growth has been running through a channel this ratio does not measure at all: secondhand tonnage. An orderbook counts steel that has not been cut. It does not count ships somebody bought last Tuesday.

Let us start with the number that is easiest to get wrong, which is 480,000. It is twenty ships of 24,000 teu, and against the 34.8m teu fleet implied by the same note it is 1.4% of world capacity. Against MSC's own 7.4m teu it is 6.5%. Neither percentage is frightening. Convert it into something with a shape in it, because teu is a unit that hides shape and shape is what decides which trade lane feels it. Assume an Asia-North Europe weekly string needs about eleven hulls to hold a weekly departure on a round trip of roughly eleven weeks; that assumption is mine, not Linerlytica's, though it is the usual shape of that trade.

Twenty hulls is a shade under two full strings. On the trans-Pacific, where a nine-week rotation needs nine ships, twenty hulls is about 2.2 weekly loops. Two services, delivered over three years. That is what the largest single order of this cycle amounts to, and it is the right way to hold it in your head: not 480,000 teu, but two weekly loops appearing slowly.

Why now, and why at this yard. Three things have to be true at once for an order this shape to get signed, and all three are true at the moment. Somebody has to have the cash, and the carriers that earned extraordinary profits in 2021 and 2022 still have it while the rest of the field has normalised. Somebody has to have berths, and a yard that was not building containerships at this scale a few years ago is where the late-2020s slots are, because the established builders are working through the book everyone placed during the last boom. I do not have that yard's backlog in front of me and neither does the note, so treat the motive as the obvious inference rather than a reported fact; the point stands either way, which is that a 62-hull programme of 20,000-plus teu ships at one yard is not a normal customer relationship, it is an anchor tenancy.

And there has to be a reason to want 24,000 rather than two lots of 12,000, and that reason is physics: a bigger hull moves a box for less fuel on the same route, and fuel is the largest controllable cost in running it. Slot cost per box is the entire argument for the megamax, and it is an argument about the carrier's unit cost, not about your rate.

Sixty-two hulls above 20,000 teu at a single yard deserves its own paragraph, because the risk in that number is operational rather than commercial. If the 42 hulls before this one average somewhere between 21,000 and 22,000 teu, the yard programme stands at roughly 1.36m to 1.4m teu, and that arithmetic is mine with the size band assumed. Against MSC's 2.94m teu orderbook, that is close to half of the largest carrier's newbuild programme riding on one yard's ability to deliver. Yards slip. New yards slip more. A yard that has never built at this scale slips in ways that are hard to predict, because the learning curve shows up around the fifth and fifteenth hull rather than in the contract.

For cargo that cuts both ways: a slipping yard means less capacity arriving on time, which supports rates, and it also means the programme gets backfilled with chartered tonnage, which is expensive tonnage that somebody's rate has to pay for. Yard concentration is the single item in this note most likely to turn a tidy delivery table into a mess.

Who feels it, and when, splits along a line that has nothing to do with size. Mainlane shippers, the Asia to North Europe and Asia to US West Coast trades, feel it last and feel least per box, because that is where the new hulls get deployed and where carriers protect the schedule. The shippers who feel it first are on the trades nobody writes headlines about: West Africa, the East Coast of South America, the Gulf, intra-Asia, the secondary Mediterranean. That is where displaced tonnage goes. An exporter moving boxes to Lagos or Santos stands to gain more from a Hengli delivery than an importer moving boxes to Rotterdam, and will get it earlier, because the ship that turns up on their lane is not new, it is the ship the new one replaced.

Importers and exporters also sit differently inside the same company. On FOB terms the freight is not your line, so a 2028 glut is a talking point your buyer will use against you on landed cost long before it is a discount you can book. On CIF, or on freight you control, it is your line, and it arrives as service before it arrives as price.

Then the timing, which is where this stops being a shipping story and becomes a contracting story. The note says plainly that none of it lands before 2028. Set that against the calendar of an annual contract. A 2027 deal signed this winter covers a year in which not one of these ships floats. The supply that is already priced into 2027 talks is therefore not supply at all: it is an expectation about 2028 being brought forward into a 2027 negotiation, and expectations cost the person who brings them nothing. Carriers will run it in two directions at once and you should watch for both. As a reason not to cut 2027 rates, because the fleet has to be paid for. And as a reason to accept a longer term, because the soft year is coming and they would rather you were locked into it when it arrives. Both arguments ask you to pay in 2027 for something that may or may not exist in 2029.

Here is the part I have not seen in the coverage, and it is the one that decides what you actually pay. A newbuild does not add capacity to your lane. It adds capacity to the lane it is put on, and then it pushes whatever it replaced down a size class, and that ship pushes another down again. Tonnage cascades. A 24,000 teu hull entering Asia-North Europe does not compete for your boxes. The 13,000 teu ship it displaces leaves the mainlane and turns up on West Africa or the East Coast of South America or the Gulf, and the 8,000 teu ship displaced from there turns up somewhere smaller still.

If you are a mid-size importer or exporter, the surplus that reaches you is not a new ship with an efficiency story attached. It is a fifteen-year-old hull looking for employment, arriving on a clock set by somebody else's replacement decision rather than by this yard's delivery schedule. That is why 46.5% is a poor forecasting tool for a single trade lane, and why the honest answer to when the glut reaches your lane is later than the delivery table, through a channel nobody can name in advance.

The second half of the same point is about what kind of object 46.5% is. It is not a forecast. It is paperwork. Shipbuilding contracts turn into steel at a rate below 100% over any multi-year window, because cancellations, conversions, yard failures and ordinary slippage all take a slice, and I am not going to invent a percentage for you, since neither the note nor I have delivery-realisation data for this specific book. Directionally, and this is the part that matters across a negotiating table, 16.2m teu is a ceiling rather than a central estimate. Every carrier sitting opposite you in a 2027 renewal is presenting the full 16.2m as though it were certain, because certainty about a glut is worth money to the seller of a two-year contract. They want you to believe 2027 is the last calm year. You should believe nothing until the steel is cut.

Now the arithmetic, and every input stays inside the note. MSC operates 7.4m teu and has 2.94m teu on order, so the orderbook is 39.7% of the operating fleet. Assume all of it delivers and MSC scraps nothing: deliberately extreme, and the right place to start, because you want the best case for cargo and the worst case for the carrier. The fleet goes to 10.34m teu, up 39.7%. Spread across four delivery years from 2028, that is about 8.7% compound fleet growth a year. Container demand does not grow at 8.7% a year; assume 3.5%, and that assumption is mine rather than a sourced forecast.

By the end of the delivery window MSC is carrying roughly 20% more nominal capacity than its volume requires. Something absorbs that gap and there are exactly four candidates: take share from competitors, slow the fleet down, scrap harder, or cut price to pull volume. They are listed in the order carriers reach for them, because the first three are cheap and the fourth is expensive. So the surplus does not land on your invoice as a discount in 2028. It lands as slower transits and more blank sailings first, and as price only when the first three levers are exhausted.

Make it yours. Assume you move 40 FEU a month on a lane where MSC is one of your carriers. That is 480 boxes a year, and every hundred dollars a box is 48,000 dollars a year. If the eventual price adjustment is 12%, and that number is mine, taken from the shape of the gap above rather than from any forecast, the annual figure is 0.12 times your all-in rate times 480. At 3,000 dollars all-in that is about 173,000 dollars a year; at 4,000 it is about 230,000. Those two levels are placeholders so you can see the shape of the formula, not quotations, so write your own rate into the line. The number that matters for planning is not the saving. It is the date. The note says none of it floats before 2028, so none of this belongs in a 2027 budget, and anyone who puts the 2028 glut into a 2027 business case is spending money that has not been earned.

One more, on the single order rather than the whole book, because it shows how much of the headline is noise. Twenty hulls at 24,000 teu. Assume each works an Asia-North Europe string and turns 6.5 round trips a year, my assumption. That is 156,000 teu of nominal slot supply per hull per year, or 3.12m teu a year across the twenty. Nominal is doing a lot of work in that sentence. A 24,000 teu ship does not sail full in both directions on every leg, so at 85% on the headhaul and materially less coming back, the effective addition is nearer 2.3m to 2.5m teu a year. Asia-North Europe is a trade measured in tens of millions of teu a year. This order, standing alone, is a single-digit share of one lane, delivered over three years. Large, and not the thing that breaks a market. The 16.2m teu is the thing that could break a market, and only if it arrives.

What would turn this from a glut into a wash, and there are three live candidates. The clearest is the gap between nominal and effective capacity. Orderbook ratios measure steel, while supply is steel times speed times routing. Slow steaming and blank sailings are the two throttles a carrier pulls without asking permission, and the industry has used both before to absorb a surplus rather than price it away. If 2028 turns soft and the response is the 2015 and 2016 response, lay up, slow down, scrap, then 46.5% never reaches the water as 46.5% of effective supply, and the relief you were shown turns up as two extra days of transit instead of a lower rate.

The second is routing. Every ship taking the long way round today is absorbing tonnage that a normal routing would hand straight back, and I will not put a number on that release because I do not have one I trust; the direction is what matters, and the direction is that a normalisation of the main east-west shortcut adds effective capacity without a single new hull, in the same window as the first of these deliveries. The third is scrapping. If demolition runs hot through 2028 to 2030, this orderbook is replacing tonnage rather than adding to it, and the oversupply story collapses into a fleet-renewal story. Any one of the three is enough. All three are live.

There is also a real argument that 46.5% is not comparable to 2009 at all, and it cuts the other way. The comparison invites you to expect a repeat of what followed the last peak orderbook, which was not kind to carriers. The fleet today is concentrated in hulls large enough that a single delivery moves a lane, and in the hands of operators who have spent five years demonstrating that they will manage supply rather than defend market share. Both facts argue for a softer landing than the ratio implies, and for the pain, if it comes, being uneven, landing on the trades where cascaded tonnage washes up rather than on the trades in the headlines.

Where that leaves anyone signing a contract is straightforward, and it is mostly about term length rather than rate. Do not accept a two-year term at 2027 rates against supply that arrives in 2028 or later; a twelve-month deal with a re-opener tied to actual deliveries, or a rate review at the half-year, is worth more than any discount you can win on the base rate. Do not consolidate your book with the carrier holding the largest orderbook on the theory that it will have the cheapest slots, because that is concentration risk bought with a story. Do not treat a lower rate as a better outcome when the carrier's first lever is speed, since a 2028 rate cut delivered alongside a two-day slower transit can cost you more in inventory than it saves in freight, and you should price the transit as well as the box. And keep the 2028 glut out of a 2027 business case, because the ships are still paper.

Forty-six point five per cent of the fleet is on order, none of it floats before 2028, and the industry would like you to sign a two-year deal before you notice the distance between those two sentences. Lock 2027 for twelve months, not twenty-four, and get the re-opener in writing now, while the ships are still drawings.

↑ Back to the key points

— By Vivian Zhao