US containerized import volumes look likely to bottom out in March 2027 rather than April, an unusually early trough that the current forecast run does not yet show. The pattern across three independent signals observed between September 9 and September 30, 2026 points to a first quarter that lands near 5.9–6.3 million TEU, with March printing below 1.95 million TEU and below both the month before it and the month after it. That would break the 2026 shape, in which the low came in April at 2.05 million TEU. The mechanism is not a demand collapse: it is a summer of tariff-driven pull-forward meeting an unusually early Lunar New Year on February 6, 2027.
In short
- The prediction: US containerized imports likely trough in March 2027, not April, with March below 1.95 million TEU and Q1 2027 landing in a 5.9–6.3 million TEU band. Base case confidence roughly 55%.
- Signal 1: The NRF and Hackett Associates Global Port Tracker revised its near months sharply up and its forward months down between its August 7 and September 9, 2026 releases, the classic signature of borrowed volume.
- Signal 2: Drewry’s cancelled sailings tracker jumped from 6% to 11% of planned capacity in a single week in September 2026, with transpacific eastbound accounting for 52% of all blanks.
- Signal 3: The Census Bureau’s advance report for August 2026, published September 30, put retail inventories up 4.8% and wholesale inventories up 6.6% year over year, so the pulled-forward goods are already landed and paid for.
- Checkpoints: the October and December 2026 Global Port Tracker releases, Lunar New Year on February 6, 2027, and the May 2027 release that reports March actuals.
Why this matters now
Import volume is the single best leading indicator retail planners have for the cost of moving a container, the availability of warehouse space, and the bargaining position of a buyer in a freight tender. It leads landed cost by roughly one quarter and leads promotional depth by roughly two. Getting the shape of the trough right, not just its existence, is what separates a useful plan from a reactive one. A March trough and an April trough imply very different tender calendars.
The 2026 cycle has been exceptionally hard to read because policy, not consumption, has driven the timing. Section 122 global tariffs of 10% expired on July 23, 2026, and Section 301 tariffs of 10–12.5% took effect the following day, which compressed a season’s worth of ordering into a few weeks. The result was an extended peak that forecasters repeatedly underestimated, as our coverage of the September import forecast revision to 2.31 million TEU set out at the time. Policy-driven peaks are followed by policy-driven holes.
There is a second reason the timing question is live right now. The transpacific service contract season opens in earnest in the first quarter, with most agreements settling between February and April for a May 1 start. Whichever spot market a buyer is looking at when they sign tends to anchor the negotiation, so a trough that lands in March rather than April changes the reference price for a full contract year. That is a durable cost consequence, not a one-quarter trading effect.
The question for the next two quarters is therefore not whether the hole arrives but when it lands and how deep it goes. The signals below point to an earlier and shallower hole than the consensus shape implies. That is a more useful conclusion than a simple bearish call, because it changes the dates on which capacity is cheapest.
Signal 1: the forecast revisions moved in opposite directions
The Global Port Tracker, published monthly by the National Retail Federation with Hackett Associates, is the closest thing US retail has to a consensus import forecast. Its month-to-month revisions are more informative than its levels, because the revisions reveal what the model is learning. Between the August 7 and September 9, 2026 releases, the near months and the forward months moved in opposite directions.
| Month | August 7, 2026 forecast | September 9, 2026 forecast | Revision |
|---|---|---|---|
| August 2026 | 2.22m TEU (-4.2%) | 2.29m TEU (-1.3%) | Up |
| September 2026 | 2.16m TEU (+2.8%) | 2.31m TEU (+9.6%) | Up sharply |
| October 2026 | 2.13m TEU (+2.7%) | 2.11m TEU (+1.7%) | Down |
| November 2026 | 2.03m TEU (+0.3%) | 2.00m TEU (-0.9%) | Down |
| December 2026 | 2.06m TEU (+2.5%) | 2.03m TEU (+1.1%) | Down |
| January 2027 | Not published | 2.09m TEU (-1.0%) | First print, negative |
September was revised up by 150,000 TEU in a single month, which moved it from a +2.8% year-over-year gain to +9.6%. Over the same revision, October, November and December were each marked down, and November flipped from a modest gain to a year-over-year decline. Jonathan Gold of the NRF described the surprise directly in the September release: “We thought the peak season would be mostly behind us by now, but that’s not the case.” Ben Hackett added that “imports have remained buoyant over the past three months despite several hurdles.”
The first published figure for January 2027 is the one that matters most here. At 2.09 million TEU it carries a year-over-year decline of 1.0%, and it does so against a January 2026 base of roughly 2.11 million TEU that was itself down 5.2% on the prior year. A forecast that prints negative against an already depressed comparison is a model saying the near-term level is structurally lower, not just calendar-shifted. That is a meaningful tell, and it is the first forward month the Tracker has published for the new year.
Read as a whole, the revision pattern is the textbook signature of pull-forward: volume is recognised in the current quarter and removed from the next. The Tracker also lifted its full-year 2026 total to 25.7 million TEU, up 1% on 2025’s 25.4 million. Full-year growth of 1% with a +9.6% September is arithmetically a statement that the back end of the year is weak.
Signal 2: carriers are cutting capacity faster than demand is falling
Carrier behaviour is a useful independent read because carriers see booking curves several weeks before volumes land. Drewry’s cancelled sailings tracker dated September 11, 2026 recorded 79 of 721 planned sailings cancelled between September 14 and October 18, an 11% blank rate. The tracker published one week earlier, on September 4, showed 47 cancellations against 729 planned sailings for weeks 37–41, a 6% rate. For weeks 38–41 specifically, announced blank sailings went from 39 to 70 in a single week.
The lane split is the detail that carries the signal. Transpacific eastbound accounted for 52% of all tracked cancellations, against 33% for Asia to North Europe and the Mediterranean combined and 15% for the transatlantic. The Pacific Southwest took the heaviest cut, with 29 blank sailings among the 78 scheduled between weeks 38–43, removing roughly 32% of capacity on the affected services.
| Measure | Week of September 4, 2026 | Week of September 11, 2026 |
|---|---|---|
| Cancelled sailings | 47 of 729 | 79 of 721 |
| Blank rate | 6% | 11% |
| Window covered | Weeks 37–41 | September 14 to October 18 |
| Transpacific share of blanks | Not disclosed | 52% |
Some of this is ordinary Golden Week management, since China’s National Day holiday ran from October 1 to 7, 2026. But the scale is instructive in both directions. Just under 13% of total planned October capacity on Asia to US and Europe services had been blanked, against nearly 18% of transpacific capacity and 21% of Asia to Europe capacity during the equivalent period a year earlier.
Spot rates held up while that capacity came out, which is the point. On September 10, 2026 Drewry put Shanghai to Los Angeles at $7,352 per 40ft container, up 2% week on week, Shanghai to New York at $9,726, up 1%, and its World Container Index composite at $4,476, flat. Asia to Europe moved the other way, with Shanghai to Rotterdam at $3,997, down 2%, and Shanghai to Genoa at $4,216, down 3%. Rates that hold only because supply is being withdrawn describe a market where carriers already expect the demand side to give way, a dynamic we examined when the transpacific rate spike was set to unwind before mid-October.
Port-level data supports the same reading from a different angle. The Port of Los Angeles handled 955,907 TEU in August 2026, roughly flat year over year and about 6% above its five-year average, with year-to-date volume just over 7 million TEU, up 1.5%. Executive Director Gene Seroka noted on September 14 that peak season imports traditionally subside at the port by the second week of November, whereas this year the lion’s share was expected to be in by the end of September. A peak that finishes six weeks early is a peak that borrowed from somewhere.
Signal 3: the pull-forward is already landed and paid for
The third signal comes from government statistics rather than the freight market, which makes it genuinely independent of the first two. The Census Bureau’s advance economic indicators report for August 2026, released on September 30, put advance retail inventories at $881.6 billion, up 0.3% on July and up 4.8% year over year. Advance wholesale inventories came in at $965.7 billion, up 0.7% on the month and 6.6% on the year. The merchant wholesaler inventories-to-sales ratio stood at 1.20 in July.
The trade figures in the same release tell the other half of the story. August goods imports were $336.1 billion, up $17.4 billion on July, and the goods deficit widened by $13.7 billion to $132.6 billion. Those are the containers that arrived during the extended peak, and they are now sitting in distribution centres rather than on the water. Inventory that is already in the building does not need to be reordered.
Wholesale inventory growth of 6.6% is running well ahead of retail inventory growth of 4.8%, which is the position you would expect if importers and distributors absorbed most of the tariff-driven buying. It also means the destocking, when it comes, likely starts upstream. That sequencing matters, because upstream destocking shows up in import volumes before it shows up in store-level markdowns, and it compounds the point we made about inventory outrunning sales at the Q3 retail earnings round.
The counterpoint deserves stating plainly: elevated inventory is only bearish for imports if demand does not absorb it. Hackett’s own commentary has repeatedly noted that consumer spending held up better than cost-of-living pressure implied. The signal here is directional rather than decisive, and it is weighted accordingly in the scenarios below.
What the pattern suggests
Put the three signals side by side and they converge on timing rather than on magnitude. None of them argues for a collapse. All of them argue that the volume that would ordinarily arrive in the first quarter of 2027 has already arrived.
| Signal | Source type | Date observed | What it implies | Strength |
|---|---|---|---|---|
| Global Port Tracker revisions reversed direction | Industry forecast | August 7 and September 9, 2026 | Volume recognised early, removed from forward months | High |
| Blank rate 6% to 11% in one week, 52% transpacific | Carrier capacity data | September 11, 2026 | Carriers pricing a soft booking curve into Q4 and Q1 | High |
| Retail inventories +4.8%, wholesale +6.6% year over year | Government statistics | September 30, 2026 | Reorder need suppressed into early 2027 | Medium |
| Peak cargo in by end of September, not mid-November | Port operator commentary | September 14, 2026 | Seasonal calendar shifted roughly six weeks earlier | Medium |
The 2026 comparison base is the complication, and it cuts against the prediction. Q1 2026 was itself a weak quarter: January came in near 2.11 million TEU, down 5.2%, February reported 1.95 million TEU, down 4.2%, and March was forecast in a 1.89–1.97 million TEU band, down as much as 12%. On those figures Q1 2026 landed close to 6.0 million TEU. Against a base that soft, modest normalisation would produce positive year-over-year prints even inside a genuine trough.
Prior precedents: how pull-forwards have resolved
Tariff-driven import surges are not new, and the resolution pattern has been consistent enough to be useful. In each prior episode the surge was recognised late by forecasters, the subsequent hole arrived faster than expected, and the depth of the hole was smaller than the height of the surge. The asymmetry matters: a +9.6% month is rarely followed by a -9.6% month, because some of the pulled-forward volume represents genuine incremental demand rather than pure displacement.
| Episode | Trigger | Surge behaviour | Resolution |
|---|---|---|---|
| Late 2018 | Section 301 list expansion scheduled for the new year | Heavy Q4 front-loading into US ports | Visibly softer first half the following year |
| 2021 into 2022 | Pandemic restocking and congestion hedging | Sustained record arrivals | Extended destocking once inventory landed |
| 2025 | Tariff effective dates through the spring | Front-loading ahead of announced dates | Full-year 2025 finished at 25.4m TEU, down 0.4% on 2024, with December 2025 at 1.99m TEU, down 6.6% |
| 2026 | Section 122 expiry on July 23 and Section 301 at 10–12.5% from July 24 | Peak extended into September at 2.31m TEU, up 9.6% | Predicted: early trough in March 2027 |
The 2025 outturn is the closest analogue, and it is instructive because the year ended marginally negative despite a strong middle. December 2025 printed 1.99 million TEU, down 6.6% year over year, and full-year volume landed at 25.4 million TEU against 25.5 million in 2024. A year of policy-driven timing noise still resolved to roughly flat annual volume.
Applying the same logic to 2027 produces a modest rather than dramatic forecast. If 2026 finishes near the Tracker’s 25.7 million TEU and the pull-forward displaced roughly one month of volume, 2027 would plausibly land flat to slightly down, with the distortion concentrated in the first quarter. That is consistent with a Q1 2027 in the 5.9–6.3 million TEU band rather than a step-change lower.
This is why the prediction is framed on shape and level rather than on year-over-year direction. The claim is that March 2027 prints the quarter’s low, below 1.95 million TEU, and below both February and April 2027. A secondary claim is that the quarter totals 5.9–6.3 million TEU, which would be flat to modestly up on 2026 and well short of the growth rate implied by a +9.6% September.
Wider context: an early Lunar New Year moves the hole forward
The calendar is the mechanism that makes a March trough more likely than an April one. Lunar New Year 2027 falls on Saturday, February 6, with the official public holiday running February 4 to 12. Chinese factories typically slow 2–4 weeks before the date and take a further 2–4 weeks to return to full capacity, so the complete shutdown cluster sits roughly between January 30 and February 14, 2027.
In 2026 the holiday fell on February 17, eleven days later. Allowing for roughly three to four weeks of transit and port handling, a February 17 holiday pushed the production hole into US arrivals during late March and April, which is consistent with April 2026 reporting 2.05 million TEU, down 7.3% year over year and the softest month of the first half. An eleven-day earlier holiday in 2027 should move that hole eleven days earlier too, into the back half of March.
| Cycle | Lunar New Year date | Shutdown cluster | Implied US arrival trough | Observed or expected |
|---|---|---|---|---|
| 2026 | February 17, 2026 | Early to mid-February 2026 | Late March to April 2026 | April 2026 at 2.05m TEU (-7.3%) |
| 2027 | February 6, 2027 | January 30 to February 14, 2027 | Second half of March 2027 | Predicted: March 2027 below 1.95m TEU |
The calendar effect also makes February 2027 look deceptively healthy. Arrivals that month reflect pre-holiday loadings from early January, which an earlier New Year tends to concentrate rather than reduce. A reasonably strong February followed by a sharply weaker March is exactly the signature an early holiday produces, and it is easy to misread as a sudden demand event.
Policy adds a second timing variable that could reinforce or disrupt the pattern. The pause on US port fees for Chinese-built and Chinese-operated vessels is a live date, and we covered the trade groups pressing USTR before the November 9 expiry. A reimposition would raise transpacific slot costs just as carriers are already managing capacity down, which would likely deepen the first-quarter trough rather than offset it.
Implications for retailers, brands, importers and investors
For importers, the most actionable consequence is the freight tender calendar. If the trough lands in March rather than April, the softest spot market of 2027 likely arrives before the transpacific service contract season concludes, which strengthens the shipper’s hand in negotiations for contracts effective May 1. Buyers who benchmark against April spot rates would be anchoring on a market that has already begun to recover.
For retail planners, the risk is a second wave of inbound ordering in Q4 2026 on top of inventory that is already elevated. With wholesale inventories 6.6% above a year earlier, the expensive error in this cycle is likely overbuying rather than stocking out. The balance of evidence favours holding open-to-buy rather than committing it early.
For warehousing and 3PL buyers, elevated landed inventory suggests short-term space demand holds into the winter even as import volumes fall. Occupancy and import volume decouple during a destocking phase, because the goods are in the building for longer rather than passing through. Contract renewals priced off import forecasts alone would likely misread that.
For investors, the read-through is to carriers and to the import-heavy discount and home categories. Carriers that can hold rates only by withdrawing sailings have limited operating leverage if volumes fall further, and the rate discipline that held the World Container Index composite at $4,476 in September is more fragile than the flat print suggests. A useful primer on how these rate cycles transmit into landed cost is our explainer on ocean freight rates and the Red Sea effect on shipping costs.
Scenarios and probabilities
| Scenario | What happens | Probability |
|---|---|---|
| Base: early trough | March 2027 prints the quarter low below 1.95m TEU; Q1 totals 5.9–6.3m TEU | 55% |
| Rebound | Restocking and resilient demand lift Q1 above 6.3m TEU; April remains the low | 20% |
| Deeper bust | Tariff escalation or a consumer step-down takes Q1 below 5.9m TEU with multiple months under 1.9m | 15% |
| Unscoreable | Methodology change, reporting gap or a disruption event that breaks comparability | 10% |
Checkpoints a reader can verify
- The October 2026 Global Port Tracker release: does it extend the forecast into February 2027, and are the Q1 months revised down?
- November 9, 2026: whether the pause on Chinese vessel port fees lapses or is extended.
- The November or December 2026 release that first frames the full first half of 2027: a figure at or below 12.7 million TEU would support the thesis.
- February 6, 2027: Lunar New Year, and the blank sailing programme announced for the four weeks following it.
- The May 2027 release reporting March actuals: the single decisive datapoint.
Caveats: what could go wrong
The strongest counter-argument is the comparison base. Q1 2026 was already a depressed quarter, with January down 5.2% and March forecast down as much as 12%, so the bar for year-over-year growth in Q1 2027 is low. A reader who scores this prediction on year-over-year direction alone would likely find the thesis wrong even if the shape is right, which is precisely why the claim is anchored on absolute level and month ordering.
The second counter-argument is forecast bias. The February 9, 2026 Global Port Tracker put first-half 2026 at 12.27 million TEU, down 2%, and the half actually landed at 12.7 million TEU, up 1.1%, a miss of roughly 3.5%. The model has been persistently too pessimistic through this tariff cycle. If that bias persists, forward months published as declines could again be revised up.
Third, pull-forward is not a conserved quantity. Firms that accelerated purchases in July and August may simply have shifted a portion of annual volume rather than borrowing it one-for-one, particularly where tariffs made inventory a cheaper asset to hold than to reorder. In that reading the first quarter softens without producing a distinct trough.
Fourth, policy could manufacture a second rush. Any new tariff effective date landing in Q1 or Q2 2027 would likely trigger the same compression that produced the 2026 peak, turning March from a trough into a scramble. The tariff calendar has been the dominant variable all year and there is no reason to assume it stops being so.
Fifth, capacity management can mask the volume picture. Carriers withdrawing 32% of capacity on Pacific Southwest services can keep rates firm and utilisation high while underlying volume falls, which makes the freight market a lagging rather than leading confirmation. A reader watching rates alone could reasonably conclude nothing has happened.
Finally, the Census inventory data is advance and subject to revision, and the inventories-to-sales ratio of 1.20 is not extreme by historical standards. The primary release schedule is published by the Census Bureau and is worth checking directly rather than through secondary coverage. Readers can follow the advance economic indicators and the full monthly trade reports on the Census Bureau’s economic indicators page.
Frequently asked questions
What exactly is being predicted, and how would a reader score it?
The primary claim is that US containerized imports at the ports covered by the Global Port Tracker trough in March 2027, with March printing below 1.95 million TEU and below both February and April 2027. The secondary claim is that Q1 2027 totals 5.9–6.3 million TEU. Both are scoreable from the NRF and Hackett monthly releases, with March actuals appearing in the May 2027 report.
Why March rather than April, given that April was the 2026 low?
Because Lunar New Year falls eleven days earlier in 2027, on February 6 rather than February 17. The Chinese factory shutdown cluster moves to roughly January 30 through February 14, and with three to four weeks of transit the resulting production hole reaches US ports in the second half of March rather than in April. The calendar shift is the single clearest mechanism in the thesis.
Could imports simply keep growing instead?
They could, and that is the 20% rebound scenario. Consumer spending has repeatedly outperformed expectations through 2026, and the Global Port Tracker’s own first-half forecast missed to the downside by roughly 3.5%. If demand absorbs the elevated inventory faster than expected, restocking could begin in Q1 rather than Q2.
Is this a demand call or a timing call?
It is primarily a timing call. The thesis does not require consumer weakness, only that the goods ordered in the summer of 2026 displace orders that would otherwise have been placed for early 2027 arrival. That displacement is visible in the forecast revisions and in the inventory data independently of any view on consumption.
What would falsify the prediction quickly?
An October or December 2026 Global Port Tracker release that revises the Q1 2027 months upward rather than downward would weaken it materially. A first published H1 2027 forecast above 12.7 million TEU would weaken it further. A March 2027 actual above 2.0 million TEU would falsify the primary claim outright.
Does the 11% blank sailing rate really mean anything, or is it just Golden Week?
Golden Week explains the level but not the acceleration. The jump from 6% to 11% in a single week, with announced blanks for weeks 38–41 going from 39 to 70, reflects decisions taken after bookings were visible. It is also notable that the blanked share of October capacity, just under 13%, sat below the prior year’s roughly 18% transpacific figure, so carriers were cutting into a softer baseline.
How does the tariff calendar interact with this?
Tariff dates have driven the timing of every import surge in 2026, beginning with the expiry of Section 122 global tariffs on July 23 and the arrival of Section 301 tariffs of 10–12.5% the next day. The November 9, 2026 expiry of the pause on Chinese vessel port fees is the next live date. A new effective date in early 2027 would likely replace the predicted trough with another scramble.
What should an importer actually do with this?
The practical implication is to delay rather than accelerate freight commitments, and to treat the March 2027 spot market as the likely benchmark for 2027–28 contract negotiations rather than the April market. Holding open-to-buy through Q4 2026 is the corresponding merchandising stance, given wholesale inventories 6.6% above a year earlier. None of this requires betting on the trough to be worthwhile, since the cost of waiting is low in an oversupplied capacity market.
Where do these figures come from?
Monthly volume figures and forecasts come from the Global Port Tracker published by the National Retail Federation with Hackett Associates, in its August 7 and September 9, 2026 releases. Capacity and rate figures come from Drewry’s cancelled sailings tracker dated September 11, 2026 and its World Container Index for September 10, 2026. Inventory and trade figures come from the Census Bureau’s advance economic indicators report for August 2026, published September 30, 2026, with port-level volumes reported by the Port of Los Angeles.