Shanghai Containerized Freight Index shows trans-Pacific rates near record

The Analysis

Standing on the buyer's side, let me lay out where we actually sit before I touch a single rate number. It's early July 2026, the third quarter is just opening, and for most importers the annual trans-Pacific service contracts were negotiated back in May. The share of volume those contracts cover varies house to house, but for a typical mid-size importer it's somewhere in the low sixties percent, which means a third or more of the year's trans-Pacific cargo still rides the spot market. That spot exposure is the part that keeps me up, because spot is exactly where the next move lands.

On 3 July the Shanghai Shipping Exchange published its SCFI, and the print is not subtle. Shanghai to US West Coast sits at USD 6,630 per FEU, up 9 percent. Shanghai to US East Coast at USD 8,296 per FEU, up 12 percent. Europe at USD 3,418 per TEU, up 2 percent. The Mediterranean at USD 4,717 per TEU, up 1 percent. The exchange itself tied the gains to improving demand and what it called a healthy supply-demand balance on the U.S. lanes. Four numbers and one sentence, and the sentence is the one I'd circle in red.

The money at the end of the day comes out of my pocket. Freight is not a line the carrier carries, it flows into my landed cost, it compresses my margin, and it shows up in the price I quote my retail and wholesale customers. So before anyone tells me to monitor the situation or stay flexible, I want to take the cost structure apart and decide how the second half of 2026's cargo gets booked and how the next contract gets signed. A lot of buyers see a rate jump and instinctively wait for it to fall. This one is different, and I'll show you why.

Start with the fact slice, because the four numbers only tell the story when you read them as a set rather than one at a time. The U.S. West Coast at 6,630 dollars per FEU and the 9 percent add. The U.S. East Coast at 8,296 dollars per FEU and the 12 percent add. Europe at 3,418 dollars per TEU with a 2 percent add. The Mediterranean at 4,717 dollars per TEU with a 1 percent add. The eye goes straight to the double digits on the trans-Pacific lanes, and that's the right instinct, those are the lanes where the move is violent. But the number that isn't printed is the explanation: improving demand plus a healthy supply-demand balance. That sentence is doing more work than the percentages do, and most readers skate right past it.

On the rate and index mechanics, Vivian Zhao has already laid out why this print behaves the way it does as a signal and what the trajectory of the index tells us about the weeks ahead, so I won't re-cut her lane-by-lane reading here. My question is the one she doesn't own: once that printed number turns into a booking, what does it actually do to a procurement ledger, and what does a buyer sign differently because of it. Her lane is the rate itself; mine is what the rate does when it hits the cost sheet.

Now the causal chain, because this is precisely where buyers misprice the situation. A rate spike driven by a capacity crunch has a shape: capacity tightens, rates jump, capacity returns, rates fall. You can wait that one out, because the cause is temporary and the cure is built into the market. But a healthy supply-demand balance is a different machine. It says demand is genuinely there, real orders, real restocking, real consumption on the other end, and supply has not added a matching increment to absorb it. Nobody throttled capacity to manufacture the rate; the market simply cleared what was offered at a higher price. When the cause is demand rather than a supply squeeze, the rate carries a floor that a pure spike never has. It does not have to give back the gain just because the calendar turns a page. That distinction is the whole ballgame for a buyer.

Layer the impact onto a cost sheet. The West Coast door at 6,630 dollars per FEU is the cheaper of the two trans-Pacific entries, yet it just got 9 percent more expensive in a single weekly print. The East Coast at 8,296 is 12 percent up and now sits about 1,666 dollars above the West Coast door, and that spread between the two U.S. doors is itself a planning variable, not a fixed fact you can ignore. Europe at 3,418 per TEU and the Med at 4,717 are up only 2 and 1 percent, which on the surface reads as calm.

But TEU pricing on those lanes and FEU pricing on the U.S. lanes are not the same unit, different equipment, different load factors, different cargo profiles, so part of that calm is an artifact of the denominator. A buyer who sees the small European percentages and concludes that ocean freight is broadly fine has missed the real signal: the lanes carrying the highest-value, most time-sensitive U.S. import programs just took the heaviest hit, and those are usually the programs with the least slack to absorb it.

Transmission timing is where this turns from a July headline into a P&L event. The SCFI is a spot reference; my contract cargo moves at negotiated rates, and my spot cargo moves at whatever I book. The 9 and 12 percent print sets the tone for every booking I place from now through the fourth quarter. The real damage window is the back half of 2026 and the early weeks of 2027: the holiday build, the spring-program pre-buy that a lot of teams already place in Q4, and any replenishment pushed through before the lunar new year. So the rate number printed on 3 July is a July event for my planning calendar, but it becomes a Q3-through-Q1 cost event for my income statement. That lag is exactly why I refuse to treat it as this month's problem that next month will solve.

Here is the point the headline didn't make, and it's the one that actually bites a buyer. The headlines said rates are near a record and climbing. True, and close to useless on its own. What they did not say is that this climb is demand-led, not shortage-led, which flips the base case. The base case for a buyer reading a spike is 'it snaps back, so wait.' The base case reading a healthy balance is 'it holds here or grinds higher, so cover.' Every playbook that opens with 'wait for it to come down' is built on the wrong premise.

The practical consequence is a timing shift: the moment to lock coverage moves forward. You stop waiting for a peak to react to and instead assume the elevated plateau is the new normal for the contracting season, and you price your second-half and early-2027 programs against that plateau. That single reframe decides whether you chase a fixed quote now or keep gambling on spot, and gambling on spot against a demand-led market is how you lose the quarter without a single dramatic event.

Run the counterfactual. Suppose I read this the old way, a spike, capacity will catch up, I'll hold spot and ride it down. Fast-forward to October. The holiday cargo still has to ship, the rate did not fall, and now I'm bidding for the same tightening space as every other importer who also waited. My patience cost me the gap between a 6,630 base I could have anchored in July and whatever the market quotes in a fourth quarter the exchange itself called healthy. The saving I imagined never showed up, and the cargo moved anyway, just at a number I no longer controlled. That is the trap of mistaking a demand-led plateau for a supply-led spike: the cargo always ships, the only variable is whether you or the market set the price.

Let me bring the buyer's own inventory language into this, because the plateau changes behavior, not just cost. My inventory-to-sales ratio and my safety-stock days are the two dials I turn when freight stops being a pass-through I can ignore. If I believe rates hold, I stop holding buffer stock at the origin just in case, because every extra day of inventory sitting in a 40-foot box is freight I have already paid, multiplied by the time it sits there. The plateau argues for compressing safety-stock days on the U.S.-bound programs specifically, since those are the lanes that just took the 9 and 12 percent hits. A tighter inventory-to-sales ratio means I forecast harder and hold less, and I accept a bit more stockout risk in exchange for not paying elevated freight on stock that isn't moving. That trade is exactly the kind of call a buyer makes and a carrier never mentions.

Payment terms are the quiet second half of this. When freight is elevated, my working capital is already stretched by any pre-buy I pull forward, and the index-linked contract's payment timing starts to matter as much as its rate. If I sign a cap with a ceiling, I want the payment window to match the shipment window, not the booking window, because the whole point of locking coverage is to know my cost before I commit cash. A carrier who wants payment on booking, while the rate still floats with the index, is asking me to fund the uncertainty I just paid to remove. I push for payment on the agreed fixed-or-capped figure, dated to when the cargo actually sails. That is a smaller negotiation than the rate itself, but on a 12 percent plateau it is where the per-unit math quietly leaks.

One more lever sits with procurement, not the shipping desk: sourcing optionality. The Europe and Med lanes moved only 1 and 2 percent, so relative to the U.S. doors they just got cheaper even though they didn't move at all. For any program I can dual-source, this is the moment to shift a slice of volume to a supplier on that side, not because the absolute cost is lower, but because the spread between the lanes widened in my favor. I'm not re-shoring a category overnight; I'm using the rate spread as one input into where the next purchase order goes. The buyer who only watches the lane she ships on misses that the cheapest door this quarter is a sourcing decision, not a routing decision.

And a word on the procurement calendar, because timing is the part buyers fumble. My next annual contract gets negotiated in the spring, but the coverage decision it locks in has to be made against this July print, not against last year's calm. If I wait for the spring negotiation to see where rates are, I have already lived through two quarters of the plateau with my spot exposure uncovered. The buyer's discipline is to re-rate the long-versus-spot split the moment a print like this lands, and to carry that re-rated split into the spring talk as a settled call, not an open question. The carrier expects me to negotiate from last year's assumption; I'd rather arrive with this year's number already baked in.

The budget line is the last place this bites. Finance builds the year's landed-cost assumption off the spring contract and the trailing SCFI, so a 9 to 12 percent jump in July lands as a variance nobody budgeted. My job is to get ahead of that variance before the quarterly review, by re-baselining the second-half and early-2027 programs off the July print instead of the May contract. If I let the old assumption ride, every container I ship in the fourth quarter shows a margin miss I could have priced in back in July. A buyer who treats the index as finance's problem loses the argument in the review; a buyer who re-baselines first walks in with the number already defended.

Let me put a number on it with assumptions I'm stating out loud, so nobody confuses this model with a fact from the article. Assume one 40-foot high-cube container carries 1,200 units of a finished good, and assume the ocean freight on that box is billed at the SCFI West Coast print of 6,630 dollars per FEU. The 9 percent increase means the incremental freight on that single box is about 597 dollars, which is 6,630 times 0.09. Spread across 1,200 units, the per-unit landed-cost lift from this one print is roughly 0.50 dollars per unit. Now assume the product's prior all-in landed cost was 18 dollars per unit; this single index move pushes it to about 18.50, a 2.8 percent lift on landed cost from one weekly print.

If the same program also ships volume to the East Coast at 8,296 with a 12 percent add, the incremental per box is about 996 dollars, or 0.83 dollars per unit on the same 1,200-unit assumption. I want to be plain: these volumes and the 18-dollar base are my assumptions, not figures from the exchange or the article, and the article gives no such procurement numbers. The mechanism is the lesson, not the specific digits, a double-digit index move reaches the unit economics almost immediately, and on a thin-margin program it can erase the quarter's margin in a single booking cycle.

So what changes in the contract? The index is the spine of every lever. The most important lever is the index-linkage clause itself. If I'm signing anything that references the SCFI, I write a cap, not just a base. A base locked at the 3 July print protects me only on the upside if the rate keeps climbing; a cap with a ceiling, the agreed rate floats with SCFI but never exceeds the July print plus a fixed band, is what actually bounds my exposure. The base lock date is the second half of that clause: I anchor the reference to the 3 July print, not to a rolling average that lets the carrier re-base upward mid-season. I'm not writing the carrier's tariff here, but the principle is 'I pay the market, not the market plus a surprise,' and the surprise is what a rolling base is designed to deliver.

Another lever is the long-term-versus-spot mix. Because the base case is a plateau and not a fall, I shift weight toward committed volume. If my current contract covers, say, 60 percent of trans-Pacific needs, this print is the argument for pushing that to 75 or 80 percent on the next program, trading a little flexibility for a known ceiling. The remaining 20 to 25 percent rides spot for the volume I genuinely cannot forecast. That split is not free, committed volume means I pay even if demand softens, and that's a real risk if the consumer side cracks, but against a healthy-balance backdrop the cost of being uncovered is higher than the cost of carrying a little slack. The trade is between two bad outcomes, and the index print tells me which bad outcome is cheaper.

There's also the port and lane routing lever. The East Coast now costs about 1,666 dollars more per FEU than the West Coast. That spread is a planning instrument, not trivia. For cargo that can absorb the longer inland leg, West Coast entry plus domestic rail is the cheaper door; for genuinely time-critical East Coast delivery, I keep East Coast volume but only the slice that truly needs it. And the Europe and Med lanes, up only 1 to 2 percent, become relatively more attractive for any program that can be sourced or fulfilled from that side of the world. I'm not recommending a wholesale reroute, lead times and duty profiles differ and those costs are real, but the relative cost ranking of the four lanes just changed, and a routing table that still reflects last quarter's spread is quietly bleeding margin.

Early shipment versus waiting deserves its own quantitative frame. If I pull first-quarter 2027 pre-buy volume forward into the fourth quarter of 2026 to beat a further climb, I pay carrying cost, storage plus tied-up working capital. If I wait and the rate holds, I pay the elevated freight on every box with no offset. The decision rule is simple to state and hard to execute: pull forward only the volume whose carrying cost is less than the expected additional freight. With a 9 to 12 percent plateau risk annualized, most programs cross that line earlier than buyers like to admit, especially the slow movers that can sit in a warehouse without spoiling or going out of season. The fast movers, the ones tied to a launch date, you don't pull, you just pay, and you build the cost into the program price.

If I were the procurement lead, here is how I'd sign. I'd anchor the next contract to the 3 July SCFI as the base, cap the float above it, push committed coverage from 60 to 80 percent, route the price-sensitive volume through the West Coast door, and pull the predictable first-quarter pre-buy forward before the fourth quarter books up. I'd treat 'healthy supply-demand balance' as a sentence that means 'plan for the plateau, not the dip.' And I'd stop congratulating myself when the Europe print only moves 2 percent, because that number is the quiet one, not the safe one.

One discipline I'll keep after this: I watch the weekly SCFI print and re-rate my coverage split monthly, not yearly. The plateau thesis only holds while the exchange keeps calling the balance healthy; the day that language shifts to capacity returning, my 80 percent committed coverage becomes the wrong bet and I unwind it. A buyer who sets a contract and forgets it cedes the timing call to the market. I'd rather be the one who moves first when the language changes. That monthly re-rate is cheap insurance against a thesis the exchange itself can flip in a single print, and I've seen it flip.

The rate is a fact; what you do with the fact is the job.

↑ Back to the key points

— By Grace Yang