MSC Launches Sierra Service to Mexico with 7 Vessels, Boosting Asia-Mexico Capacity Alongside Mexicas Loop

The Analysis

The first time I had to move cargo into Mexico I learned the map the hard way. The ocean leg is the easy part. What decides whether a shipment works is which side of the country you come out on, and how many kilometres of mountain road sit between that port and the factory gate. Manzanillo leaves you on the Pacific side with a long climb up to Guadalajara and the Bajío industrial belt. Lázaro Cárdenas leaves you closer to Mexico City and the centre of the country. Ensenada is almost a different country altogether, sitting up in Baja California next to the border factories at Tijuana and Mexicali. Three doors, three completely different inland problems.

That is why the new MSC service matters more than the vessel count suggests. Sierra runs seven ships of between four thousand and nine thousand TEU on a rotation of Shanghai, Qingdao, Busan, Manzanillo and Lázaro Cárdenas before turning back to Shanghai. It sits alongside the existing Mexicas loop, eight vessels of five thousand TEU running Ningbo, Shanghai, Lázaro Cárdenas, Manzanillo, Ensenada and back to Ningbo. Two loops, fifteen ships, four Asian loading ports and three Mexican gateways between them.

Read that as a route map rather than a press release and it changes what you can promise a customer. The dual structure is not really about more boxes. It is about having more than one way into the country, and about which of those ways suits which inland destination. The trade behind it is the nearshoring story everyone is talking about, factories moving to Mexico and supply chains spreading out from China, but nothing in that story removes the fact that the components still have to cross the Pacific first. If anything it makes the crossing more important, not less.

Start with the Asian end, because that is where most people make the first mistake. Shanghai appears on both loops, so a shipper loading in Shanghai can in principle use either service. Ningbo appears only on Mexicas. Qingdao and Busan appear only on Sierra. That single difference decides how much redundancy you actually have, and it has nothing to do with who your carrier is. If your boxes are stuffed in Ningbo because that is where your supplier sits, then the new service does not give you a second option at all, unless you are willing to dray them north to Shanghai first, which is a decision about inland cost and a day or two of transit, not about which vessel you book. If you load in Qingdao, the new loop is the one that reaches you directly. If you load in Shanghai and you have a supplier who can also make Ningbo, you are the one holding the flexible hand.

Busan is worth a paragraph of its own. A Korean call on a Far East to Mexico rotation is not really about Korean cargo. It is a transhipment valve. North and north-east Asian cargo moves down to Busan and transfers onto the mainline, which means a factory in Dalian or Tianjin can reach Manzanillo through this service without ever touching a Chinese main port. That is useful, and it is also a second set of sailings you are exposed to, because the feeder leg and the mainline leg each have their own schedule. If you route through Busan, get both vessel names in writing and assume either one can move you a week.

Now the Mexican end, which is where the real decision gets made. Manzanillo is the big Pacific gateway and it serves a particular geography: Guadalajara, the Bajío belt around León, Querétaro, Guanajuato and Aguascalientes, and the automotive and appliance plants that cluster there. If your delivery address is in that belt, Manzanillo is the natural door even when it is busy, because the alternative puts you on a longer road from somewhere else. Lázaro Cárdenas sits further south in Michoacán and is meaningfully closer to Mexico City and the centre, with a rail corridor running up toward the capital that changes the arithmetic for cargo heading to the Valley of Mexico or to Querétaro.

Ensenada is small, northern and specific: it serves Baja California and the border plants at Tijuana and Mexicali, and it is the right door for a maquiladora operation rather than for a distribution centre serving the whole country. Choosing between them is not a matter of preference. It is a matter of where the truck or the train has to go afterwards.

Say it plainly: this is a road problem with a sea leg attached. I have watched people book the port with the best schedule and then lose more days on the inland leg than they saved on the water, because they picked a gateway on the wrong side of the country. Do not pick a port because it looks good on a rotation. Pick it because the road out of it goes where you are going.

The redundancy angle is the one I would spend time on. Fifteen vessels across the two loops means the service is not a single point of failure any more. If one Mexican gateway backs up, there is another one on the same carrier's network that your cargo can legally and physically move through. That is a real improvement and it is worth money, but only if you have done the work beforehand. A second gateway you have never used is not a backup, it is a theory. You need a drayage company that covers it, a rail booking habit if you are going to use rail, a customs routine that has actually been run there at least once, and a warehouse receiving window that can absorb a different arrival pattern. Build those before you need them and the second loop is worth a great deal. Build them during a congestion event and you will spend three weeks learning what everyone else already knows.

Here is the capacity arithmetic, with the assumptions stated. Sierra has seven vessels in the four thousand to nine thousand TEU range. Assume an average of around six thousand TEU, which sits inside that range without pretending to be precise, and assume a weekly rotation. A weekly service deploys roughly one vessel's worth of capacity each week, so Sierra puts about six thousand TEU of nominal capacity a week into the lane. Mexicas has eight vessels of five thousand TEU, which on the same weekly assumption is about five thousand TEU a week. Together that is around eleven thousand TEU a week of nominal capacity on the Asia to Mexico direction. Nominal is the important word: it is the vessel space, not the bookable space, and actual availability depends on the sailing schedule, on how much of the ship is committed elsewhere, and on the season. Check it against the published rotation rather than against my arithmetic, but the order of magnitude is the point.

There is a second, cleaner piece of arithmetic from the same numbers. On a weekly service the number of vessels roughly equals the number of weeks in the round trip, because one ship has to cover each week of the cycle. Seven ships on Sierra implies a round trip of about seven weeks; eight ships on Mexicas implies about eight. Those are long loops. Long loops are good for port coverage and bad for recovery speed: when a single vessel is pulled out, the hole travels for weeks rather than days. So the dual structure gives you breadth, not necessarily resilience in the short run. Two loops into three ports is much better than one loop into one port. It is not the same thing as two departures a week that you can swap between at the last minute.

Now a time calculation, because that is what actually decides whether the second port is worth using. Assume you move forty 40ft containers a month into the Mexico City region, all of them currently through Manzanillo. Assume the inland leg from Lázaro Cárdenas into that region is about one and a half days shorter than the inland leg from Manzanillo, which is an assumption about distance and road conditions rather than a published figure, so check it against your own transit data. Move half of those boxes, twenty of them, to Lázaro Cárdenas and you save twenty times one and a half, which is thirty container-days a month.

Assume an average cargo value of sixty thousand dollars per container and an annual carrying cost of nine per cent. One container-day costs roughly sixty thousand times nine per cent divided by three hundred and sixty-five, which is about fourteen dollars and eighty cents. Thirty container-days is therefore around four hundred and forty dollars a month. That is the honest size of the inventory benefit, and it is small.

The value is not in that four hundred and forty dollars. The value is in what happens on the week Manzanillo backs up. Assume a congestion event adds five days of berth and yard time at Manzanillo, which is the kind of number that port has produced before when volumes spike. Assume your monthly forty boxes are evenly spread, which is about nine or ten boxes a week. Five days of delay touches roughly seven boxes, and if those seven boxes feed a line that cannot stop, you are now looking at an expedited move or a stopped line, and both of those cost more than a year of the inventory saving. Carrying half your volume through a second gateway is insurance priced in days rather than in money. The thirty per cent of volume you keep on the second door is the spare tyre, and you do not buy a spare tyre to save fuel.

Build the time budget the way I would build it: as a list of handovers, each with its own assumed range. Loading port to vessel, say two to four days depending on cut-off and how early you gate the box. Ocean leg, taken from the published rotation and treated as an estimate rather than a promise. Discharge waiting time, assume zero to three days in a normal week and more when the yard is full. Clearance, assume one to three days and remember that the first time you clear at a new port it will sit at the slow end of that range.

Inland leg by road or rail, assume one to three days to the Bajío belt from Manzanillo and less from Lázaro Cárdenas to the centre. Add them and you have a door-to-door range of roughly three to five weeks, most of which is water. The point of writing the list down is that you can see which handover is worth attacking. In my experience it is almost never the ocean leg. It is the discharge wait and the inland booking.

When Manzanillo does back up, the detour menu is short but real. Move the cargo to Lázaro Cárdenas if the destination is the centre of the country. Move it to Ensenada if the destination is the northern border. Use rail out of Lázaro Cárdenas rather than road if road capacity is what has tightened, and book the rail space early, because a rail slot you have not reserved is not a slot. Or take the longer way round: discharge at Los Angeles or Long Beach and truck the cargo south across the border into Tijuana or Mexicali, which makes sense for northern destinations and very little sense for anything heading to the centre. Each of those moves has a cost in days and a cost in the number of relationships you have to re-point. The cheapest detour is the one you have already run once.

That northern option deserves more than the one line I gave it, because it is the one people reach for at the wrong moment. Discharging at Los Angeles or Long Beach and trucking south across the border can genuinely beat a direct call when your destination is Tijuana, Mexicali or the northern industrial parks, and when the Pacific port you would otherwise use is the one that happens to be congested. It works because the US west coast has more sailings, more frequent departures and a deeper truck market, so the ocean leg is easier to re-book.

It stops working the moment the cargo has to continue to the centre of the country, because you have then added a border crossing and a very long road to a route that already had one long road in it. It also adds two clearance events instead of one, and a border queue that does not read your schedule. Use it for the north. Do not use it as the general answer for Mexico, and do not find out about it in the week you need it.

One more practical thing, and it is the one that quietly decides whether a Mexican programme holds together: where the empty box goes afterwards. On a lane like this the import container does not always leave the country through the port it came in through. If your inland destination has no convenient empty return point, the drayage company runs a round trip and you pay for both halves, and the half with no cargo in it is the expensive one. Ask, for each gateway you are considering, where empties are actually accepted, how far that is from your delivery address, and what the driver does after he drops your box. It is an unglamorous question and it moves the inland cost structure more than most of the things people spend weeks negotiating. The same logic applies to chassis and to rail space: the binding constraint usually sits at the end of the road, not at the port gate.

And work the calendar from both ends. A Mexican plant running on tight inbound schedules has very little tolerance for a door-to-door range of three to five weeks, so the way to make this work is to build the plan backwards from the line call-off date: fix the inland leg first, then the clearance allowance, then the berth allowance, and only then pick the departure that makes the arithmetic close. Shippers who plan forwards, starting from the sailing and working outwards, are the ones who end up expediting. Assume the first two cycles through any new gateway will sit at the slow end of your assumptions, and tell the plant that in advance, rather than letting them find out on the day.

There is a wider point here that I think gets lost in the nearshoring conversation. Everyone talks about production moving to Mexico as if it shortens the supply chain. For finished goods going north into the United States, it does. For the components, the tooling, the packaging and the semi-finished sub-assemblies that feed those new Mexican plants, it does not shorten anything at all. Those still come from Asia, and now they go to a port system that has historically run tighter than the US west coast, into an inland network where road capacity, not port capacity, is usually the binding constraint. More vessels and more port calls on this lane are a response to exactly that. They are not a luxury. They are the difference between a nearshoring plan that works and one that stalls at the port gate.

The traps are predictable. Ensenada is small, and treating it as an equal-sized backup for Manzanillo is a mistake; its inland reach serves the north-west and the border, not the industrial centre. Rail out of Lázaro Cárdenas needs space booked in advance, and if you assume you can call on the day you will be disappointed. Truck capacity around Manzanillo tightens in the same weeks the port gets busy, which means the congestion costs you twice, once waiting for the box and once waiting for a truck.

Changing gateway is not a booking change, it is a full re-pointing of the inland plan: new drayage company, new appointment system, new receiving pattern at your warehouse, and a customs routine nobody on your team has run before. Free time and detention run on calendar days in most contracts, so yard congestion converts directly into equipment cost. And the arrival date on the carrier's system is a forecast; the only date that behaves like a commitment is the one in your own customer contract.

What I would do this month is straightforward. Map every Mexican delivery address you serve to a gateway, using inland geography rather than habit, and finish that mapping before the middle of July. Then put a defined share of volume on the second loop, not as a one-off experiment but as a standing rule, because the only way the second door stays usable is if your team uses it often enough to remember how. Then run one test shipment through the gateway you are least familiar with, on low-risk cargo, and record the actual door-to-door days so you have a real number instead of an assumption. Then set the trigger: if berth or yard time at your primary gateway runs beyond five days, the second gateway gets evaluated within forty-eight hours. And book inland capacity further ahead than you do today, because the road and the rail slot are the parts that fail when the port gets busy.

Two loops, fifteen ships, three Mexican doors. On paper that is capacity. On the ground it is a choice about which road you drive at the other end, and how many roads you have bothered to learn. I have crossed this country enough times to know that the port is rarely the hard part. The hard part is everything between the port and the gate, and the more doors you have actually used, the less any single one of them can hurt you.

↑ Back to the key points

— By Ömer Kaya