When to Order Wholesale Backpacks: What 53 Months of Customs Data Say About the Calendar

Rewritten 11 August 2026 — this guide has been rebuilt on a real monthly customs series rather than on a supplier's intuition about the calendar. The previous version asserted that late summer carries 30–40% of annual backpack revenue, that demand spikes 3–5x, that stock must be off the water by early July, and that ocean freight from Asia takes 30–45 days. None of those had a source, and the measured data below contradicts the demand multiple, the early-July deadline and the transit time. All corrected in the open.

Almost every backpack ordering calendar on the internet is written from the retail selling season backwards, using lead times the writer has never measured. There is a better way to do it, and it has been sitting in a public database the whole time: the European Union publishes, month by month, the customs value of every rucksack it imports from outside the bloc. Fifty-three consecutive months of it. That series tells you when containers actually land — not when a supplier thinks they should.

What follows is that series, the two other independent series that corroborate its shape, the freight prices attached to getting the timing wrong, and a worked backward plan from a target on-shelf date to a purchase-order date with every input either cited or flagged as our own assumption. Three things in it will change a buying calendar: when the containers actually arrive, how much later than you think that is, and the one piece of seasonal freight advice that was false in 2026.

1. The seasonal shape is measured, not asserted — and it survived three different shocks

The Eurostat Comext database, dataset DS-045409, carries monthly extra-EU import values at the eight-digit Combined Nomenclature level. Code CN 4202 92 91 is the backpack line: "travelling-bags, toilet bags, rucksacks and sports bags, with outer surface of textile materials." Pulling EU27 as reporter and all non-EU countries as partner, indicator VALUE_IN_EUROS, flow 1, gives an imports-only monthly series running from January 2022 to May 2026 — 53 consecutive months and four complete calendar years.

Here is what it looks like.

Three separate things are visible in that table, and only one of them is the thing everybody claims.

What is solid: in every one of the four complete years, June, July and August together took more than their flat share of the year's arrivals. A perfectly even year would put 25.0% of import value in any three months. The actual figures were 30.1%, 27.0%, 29.5% and 27.4%. Averaged across those four years, the seasonal index — each month expressed as a percentage of that year's average month — peaks in July at 118.9, with June at 112.0 and August at 111.1, and bottoms in November at 85.1 and December at 86.0.

What is less tidy, and we are printing it because it is true: the single highest month of the year sat outside the June–August block twice. In 2023 the top month was May (EUR 163.3 m) and in 2025 it was March (EUR 174.4 m). Anyone telling you imports peak in July every year without exception has not looked at 2025. The block over-indexes reliably; the individual month does not.

What is remarkable is the period the shape survived. 2022 was the tail of the container-rate boom. 2023 was a destocking year, and you can see it — total import value fell from EUR 1,888.4 m to EUR 1,607.4 m, a drop of nearly 15%. 2025 and 2026 carried tariff disruption heavy enough to distort freight pricing for months at a time. The level moved a great deal. The summer weighting did not.

What this means for you: build your calendar on the three-month June–August arrival block, not on a single peak month. A plan that depends on July specifically will be wrong three years in four — July was the single highest month only in 2024; 2022 peaked in August, 2023 in May and 2025 in March. A plan that treats June through August as the landing window is consistent with all four years of measured data.

2. Three independent series, three summer peaks — and one of them is not what it looks like

One series is a fact about one market. Three series measured by three different authorities, in three different units, are harder to dismiss. Alongside the Eurostat import figures we pulled UK trade in the same commodity code from HMRC's Overseas Trade Statistics, and monthly loaded import containers from the Port of Los Angeles.

The Port of Los Angeles series carries its own caveat, of a different kind. It is a real measurement, published monthly by the port authority itself, of loaded import containers in TEU — but it is all commodities, not backpacks. It appears here because no US series at this commodity code could be obtained at all: the Census Bureau's international-trade API requires a key, USITC DataWeb and USA Trade Online both require an authenticated account, and the UN Comtrade free preview endpoint returned compressed binary that could not be decoded. That gap is listed again at the end of this article, because it is the largest one in it.

With those two caveats stated, the pattern is the pattern. The EU import series peaks in July at 118.9. UK total trade peaks in July at 126.2, and did so in three of the four years measured; 2023 peaked in June. Port of LA loaded imports peak in August at 115.0, with July to September as the top three months. Three authorities, three units, three peaks, all inside a ten-week window.

Note also the trough. The EU index bottoms at 85.1 in November; UK total trade bottoms at 76.7 in December — a 1.65x swing from its July high. The quiet end of the calendar is quieter than the busy end is busy.

What this means for you: if your own arrival dates do not sit inside June to August, you are either doing something deliberate or you are late. Being deliberate is a legitimate answer, and section 8 sets out why. Being late is not.

3. The container clears customs one to two months before sell-through, not six

This is where the standard advice breaks, and the break is worth about two months of working capital.

The old version of this page said that back-to-school stock "must be on your shelves — not on the water — by early July." Read the measured arrival curve again: July is the peak arrival month in Europe. Containers are not late in July; July is when the market's containers land. Meanwhile the retail sell-through those containers serve happens afterwards. Using US Census Bureau Monthly Retail Trade Survey figures, not seasonally adjusted, the seasonal index for sporting-goods stores (NAICS 45111) puts August at 111.3, the second-highest month of the year, and for clothing and clothing-accessories stores (NAICS 448) puts August at 103.8, the third-highest.

Put the two curves side by side and the gap is one to two months for the summer season, not six. Arrivals index above 110 in June, July and August; sporting-goods sales index above 105 in June, July and August and peak, within that window, in August. That is a supply chain running about one month of slack between customs clearance and the shelf, and the honest description of it is that it is tight — not that it is generous.

The December retail peak is a different story and deserves its own warning. Clothing-store sales index at 157.2 in December, the largest single-month number anywhere in this article, and sporting goods at 140.7. Arrivals in the EU series, by contrast, are at their annual floor in November and December. Cross-correlating the Port of Los Angeles loaded-import index against the US clothing-store index — the two series that describe the same market — the best fit has imports leading retail by four months, with a correlation coefficient of 0.44. That is moderate, not strong, and it is our own arithmetic rather than anybody's published finding. Treat it as a hypothesis to test against your own receipts, not a rule.

What this means for you: stop planning a six-month gap between clearance and sell-through for the summer season, and stop assuming the December peak works the same way. The summer programme lands roughly a month ahead of the till. The Q4 programme, on this evidence, arrives around August — which is exactly when the summer containers are also landing. Whether that is what congests warehouses in the third quarter is a plausible reading of two curves, not something we can source.

4. The Chinese New Year advice was wrong in 2026, and it is worth knowing why

Every sourcing calendar tells you to book ahead of Chinese New Year because rates spike into the holiday. In 2026 the opposite happened, on the record, week by week.

Chinese New Year 2026 fell on 17 February. The lowest composite reading of the year to date — USD 1,919 — is the week 08 assessment, which Daily Cargo News dates 19 February and Shipping Telegraph republished on 20 February. Either way it lands after the holiday, not before it. And it was not a single odd week: the 12 February assessment of USD 1,933 was described in Drewry's own commentary as the fifth consecutive weekly decline (Daily Cargo News, 12 February 2026), and week 08 was the sixth.

Drewry said out loud what was happening. In its 15 January release, republished by AJOT, it noted that carriers faced upward pressure "due to the expected Chinese New Year factory shutdowns in mid-February". Rates fell anyway. The same thing had happened around Golden Week the previous year: the assessment for the week of 11 September 2025 recorded a thirteenth consecutive weekly decline heading into a holiday that began on 1 October, with Drewry commenting that "despite the upcoming Golden Week holidays in China, it is unlikely that these rates will be sustained without further cuts to shipping capacity."

There is a further point buried in the 2026 record. The single largest weekly rate move in the period we examined had nothing to do with a Chinese holiday at all. Freightos reported FBX01, Asia to US West Coast, rising 51% in the week of 9 June 2026 to around USD 4,800 per FEU — its sharpest weekly rise since June 2025. Freightos attributed the move to contracted shippers pulling shipments forward ahead of an 80% jump in fuel surcharges due in July, and to Red Sea diversions lengthening lead times. Deadlines move rates. Holidays move capacity.

What this means for you: treat Chinese New Year as a capacity and calendar event, and price it as a production risk rather than a freight-price risk. Book around the factory shutdown because your goods will not be made otherwise. Do not pay a premium in January on the assumption that February will be dearer — check the index for the year you are actually in.

5. Ordering late is priced, and the price is on the record

The strongest available evidence for what a late order costs is not a markdown study. It is the same container, on the same lane, assessed twenty weeks apart.

USD 4,263 per container is the number to keep. It is not a modelled figure, not an average of anything, and not a survey: it is the gap between two published assessments of the same slot, twenty weeks apart. On a mixed 40ft container of backpacks that is real money against a landed cost, and it is entirely a consequence of when the booking was made.

Two things pushed rates up over that window, and both were visible in advance. A published peak-season surcharge of USD 2,000 per 40ft took effect on transpacific eastbound from 1 June 2026, and CMA CGM announced forward FAK rates from the same date of roughly USD 4,700 per 40ft Asia–Europe and USD 5,500 to 5,700 per 40ft Asia–Mediterranean. Surcharge effective dates are published weeks ahead. A buyer who reads them is not guessing.

The fallback everybody reaches for is air, and here the honest position is that we can price the air leg but not the comparison. Freightos published its air index for the week of 29 July 2026 at USD 5.76 per kg China to North America and USD 3.84 per kg China to North Europe; by 7 August 2026 the China–US figure was USD 5.67 per kg and China–Europe USD 4.02 per kg, described as "more than 10% lower than a month prior". What nobody publishes is the chargeable weight of a container of backpacks. Backpacks are low-density and high-cube, so the ocean cost per kilogram depends entirely on your own pack-out. We are not going to print an "air is N times ocean" multiple, because computing one would require an assumption we would then be hiding inside a ratio.

What this means for you: put the freight index in the same spreadsheet cell as the unit price when you evaluate a late order. A supplier discount of two per cent on a container that now costs USD 4,263 more to move is not a saving. And if air is your contingency, get a chargeable-weight figure for your own cartons before the season, not during it.

6. The lead time your factory quotes starts later than you think it does

The old version of this page said production takes "30–60 days for standard wholesale orders; 45–75 days for custom or private-label programs once samples are approved". We could not source either range, and the figures we could source are materially longer at the custom end.

The structural point matters more than any individual number. Osgoodway's clock starts at "confirmed style selection" for ODM and "locked tech pack" for OEM. Everything before that moment — artwork iteration, material selection, colour approval, the sample rounds themselves — sits outside the quoted lead time. It is also precisely where schedules are lost, because it is the only part of the sequence controlled by the buyer rather than the factory.

Huafang's disclosure on sampling models is worth reading closely for the same reason. It documents two arrangements: one where "the sampling deposit collected before sampling will be fully refunded to the customer when they place a bulk order", and one where "the sampling fee will generally not be refunded to the customer after the sampling is completed". In the refundable model, mould and material fees charged by third parties are explicitly excluded from the refund — which is where a USD 200 to 800 mould charge lands. A free sample and a sample fee credited against bulk are not the same offer.

Then there is the factory calendar itself. Under State Council Decree No. 795, in force from 1 January 2025, Spring Festival carries four statutory days rather than three, and the decree added a rule that continuous work before or after a statutory holiday is "generally not more than 6 days" — which, with the make-up working weekend mechanism, stretches the statutory core into a contiguous block. In 2026 the published arrangement was a nine-day holiday from 15 to 23 February, with make-up working Saturdays on 14 and 28 February.

Two consequences follow that are easy to miss. Chinese New Year 2027 falls on a Saturday and 2028 on a Wednesday, and a midweek new year forces a longer make-up bridge than a weekend one — so the shutdown is not the same length every year, and the day of the week is the reason. And in 2026 the Mid-Autumn holiday and Golden Week are separate stoppages four days apart, which is arguably worse for production continuity than 2028's single merged block.

What this means for you: date your lead time from the day the specification is locked, and confirm which side of Chinese New Year the factory is counting from. If you are running an OEM programme at 90 to 120 days, an August delivery means a specification locked in the first quarter — and if that lock lands in February, the factory calendar, not the factory, will decide when it starts.

7. The longest step in the chain is not the riskiest step

Planning calendars almost always budget for the longest leg. The evidence says the longest leg is the most predictable one, and that fixed launch dates die in the short steps.

Read the reliability figure carefully, because it is usually read wrongly. 62.6% reliability does not mean vessels are on average 37% late. It means roughly 37 in every 100 vessels arrive late, and those that do are late by 5.31 days on average. The right buffer is therefore not "average transit plus a little" but something closer to the mean lateness weighted by the probability of lateness — and the sensible planning move is to take the full 5.31 days, because you cannot know in advance which sailing you have.

The carrier spread is the bigger lever, and it is one buyers rarely negotiate. In June 2026 Maersk ran at 77.1% and Wan Hai at 35.6% on the same trades — 41.5 percentage points apart. Only three carriers were above 70%; eight of the remaining nine sat between 50% and 60%. Which carrier your forwarder books is a bigger schedule variable than which route they book.

The port tail is the third trap. Los Angeles import dwell is not normally distributed — it is bimodal. Three quarters of boxes clear within four days, and then 13.9% sit for nine days or more. An average dwell figure describes almost no actual container. At the other end of the chain, delay is priced and published: the Port of Felixstowe charges no storage where units are collected within five days from commencement of discharge, then GBP 50.05 per day for a 40ft box on days six to twelve, and GBP 135.28 per day from day thirteen. A fortnight of clearance or haulage trouble beyond the free period — days six to nineteen — costs GBP 1,297.31 per 40ft container in quay rent alone on that published tariff: seven days at GBP 50.05 plus seven at GBP 135.28, before anything the carrier charges in demurrage. UK public holidays are excluded from the free-time count, which quietly lengthens free time over Christmas and Easter.

What this means for you: buy your buffer where the variance is, not where the days are. Ask for the carrier name and its most recent schedule-reliability figure before you approve a booking; ask your broker which port of entry they intend to clear through and what its inspection rate looks like; and price five days of quay time into the plan rather than hoping for none.

8. The cheapest, emptiest months to land a container are the months nobody books

The trough in the arrival data is as informative as the peak, and almost nobody plans around it.

If some part of your range is not seasonal — plain commuter packs, replenishment lines, corporate stock — then the fourth quarter is a structurally quiet window in the arrival data, and the freight index bottomed in the first quarter in 2026. Both of those are measurements. The step from "arrivals are low and rates were low" to "you will get a better slot and a cheaper rate" is an inference and we are labelling it as one, because the causal chain runs through carrier capacity decisions we cannot observe.

The publication lag deserves its own line in your process. Both official series run about ten to eleven weeks behind. You cannot use Eurostat or HMRC to react inside a season — by the time June's arrivals are published, the season is over. They are planning inputs, checked once a year, not a dashboard. Port of Los Angeles is much faster, publishing the prior month during the second half of the following month, which makes it the only one of the three worth watching in-flight.

What this means for you: split your calendar in two. Seasonal lines follow the June–August arrival window and everything in this article. Non-seasonal lines should deliberately avoid it, and the fourth quarter is where the arrival data says there is room. If you are buying stock-priced catalogue items rather than a bespoke programme, that flexibility is worth more than a unit-price negotiation — you can see the stock-priced range in bulk deals.

A worked backward plan, itemised, with the assumptions marked

Here is the whole thing applied to one order. The target is 1 August 2027, which puts stock on shelf just ahead of the August retail index peak in the US series and inside the June–August arrival window in the EU series. The buyer is in the UK, the origin is Shanghai, the port of entry is Felixstowe, and the order is an ODM catalogue style rather than a bespoke OEM programme.

Three observations about that plan, all of which change the answer if your situation differs.

The port of entry moves the purchase-order date by two weeks. CMA CGM's FAL 8 publishes Shanghai to Rotterdam at 40 days, Hamburg 46, Antwerp 50 and Felixstowe 55; from Ningbo the same service reads 42, 48, 52 and 55, and from Yantian 44, 50, 52 and 58. Substituting Rotterdam for Felixstowe in the plan above pulls the ocean leg from 55 days to 40 and moves the purchase order from 11 March to 26 March. ONE's FE3 service, effective from June 2026, publishes about 48 days Qingdao to Felixstowe, verifiable from the two ETAs it gives for the first voyage — Qingdao 11 June 2026, Felixstowe 29 July 2026.

OEM moves it by a month or more. Swap Osgoodway's 45–60 day ODM figure for their 90–120 day OEM figure and the purchase order moves from 11 March 2027 back to roughly 10 January 2027 at the 120-day end — which lands the entire production start inside the Spring Festival window. That is the single most common way a bespoke programme misses an August delivery.

Every Asia to North Europe transit we could find published is a Cape of Good Hope routing. ONE's FE1 and FE3 rotations state it explicitly, and CMA CGM's FAL 8 day counts are consistent with it. We did not audit every carrier. This is why 40 to 58 days is now normal and why any article still quoting 30 to 45 days — including the previous version of this one — is structurally wrong rather than merely stale.

The checklist: what to ask a supplier, and what to put in your own calendar

The markdown statistic we deleted, and why there is nothing to replace it with

The previous version of this page implied a cost to missing the season — unsold peak stock, markdowns, tied-up open-to-buy. That instinct is sound. The number that usually accompanies it is not, and we could not find a usable one anywhere.

We searched for a formally published retail statistic on the markdown cost of a missed selling season. The only item located in the formal literature was McGoldrick, P.J. and Betts, E.J., "Seasonal Markdown Strategies of Apparel Retailers: Audit Evidence and Consumer Preferences", in the Proceedings of the 1997 Academy of Marketing Science Annual Conference — and only its abstract is accessible, describing "audits of pricing and mark-down activities between 1993–96" with no extractable percentage. Every quantified markdown figure we found elsewhere traced back to a software vendor's blog.

So there is no markdown number in this article. There is a freight number instead, because that one is real: USD 4,263 per 40ft container between the February and July 2026 assessments of the same Shanghai–Los Angeles slot. A defensible small number beats an impressive unsourceable one, and the same logic applies to the "30–40% of annual revenue in eight weeks" and "3–5x baseline volume" claims the previous version carried. Neither had a source. The nearest measured equivalents point the other way: June to August took 27.0% to 30.1% of annual import value depending on the year, and the within-year peak-to-trough ratio in the EU series ran 1.52x to 1.68x — nowhere near three to five times.

What this means for you: if a supplier quotes you a markdown percentage for missing a season, ask where it comes from. We looked, and the honest answer appears to be that nobody has published one you can use.

Ordering with Backper

Applying this page's own tests to us, in the same format we would want a supplier to answer them.

You can see current back-to-school silhouettes in the school collection and stock-priced items in bulk deals, or ask for a quote directly.

Related reading. For the freight, duty and clearance steps in detail, see shipping, freight and customs. For decoration lead times on a branded programme, see logo printing methods. For school-programme specification and compliance, the school backpack buying guide.

What we could not verify

• A US monthly import series at HS 4202.92 or below. This is the largest gap in the article. Three routes were tried and all failed: the Census Bureau international-trade API returned "A valid key must be included with each data API request"; USITC DataWeb and USA Trade Online both require an authenticated account; and UN Comtrade's free preview endpoint returned gzip-compressed binary that could not be decoded. Port of Los Angeles loaded imports are used as a measured substitute, but they are all-commodity, not backpacks. Anyone with a free Census API key can close this gap in an afternoon.
• A UK imports-only monthly series. HMRC's OData implementation silently drops the aggregation when the filter carries three conditions, so flow could not be separated from month and commodity. The UK figures in this article are total trade, imports plus exports, measured at 79.0% imports across the full 2000–2026 history. This is stated everywhere the series appears and it should not be republished as an import series.
• Any published retail statistic for the markdown cost of a missed selling season. Searched at length; the only formally published item found was a 1997 conference proceedings abstract with no extractable figure. Every quantified number located elsewhere traced to a vendor blog. No markdown figure appears in this article and the previous version's implied one has been deleted.
• The old article's 30–40% of annual revenue, 3–5x demand spike, 30–45 day ocean transit, 8–12 day air transit and 10–20 day nearshore road figures. None could be traced to any source. Two are contradicted by the data above. All have been removed rather than softened.
• EU import quantities in kilogrammes or units. Only VALUE_IN_EUROS was retrieved from Comext. The dataset also carries QUANTITY_IN_KG. All EU seasonality here is therefore value-based and mixes price effects with volume effects. The same caveat applies to the UK figures: requesting NetMass alongside Value reliably timed out.
• Any series for CN 4202 92 11, the plastic-sheeting-faced backpack line. Not pulled. The textile code is the dominant line but the plastic line is unquantified here, and it carries four times the UK duty.
• The EU third-country duty rate for CN 4202 92 91. The instrument is identified — Commission Implementing Regulation (EU) 2025/1926, published in the Official Journal on 31 October 2025 and applying from 1 January 2026 — but the rate line itself was not retrieved after ten distinct attempts across TARIC, EUR-Lex, Access2Markets and three national tariff browsers. No EU rate is printed. Read it off TARIC before it goes into a costing.
• The State Council public-holiday arrangement for 2027 and 2028. Neither exists yet. Only the astronomical Lunar New Year dates from the Hong Kong Observatory are citable today, plus the four statutory days fixed by Decree No. 795. The 2027 notice is expected around November 2026.
• Any Chinese official post-holiday return-to-work statistic. No provincial-government, port-authority or chamber-of-commerce release with a quantified 2026 figure was located. Maersk's "2–3 weeks before", "not to full capacity until mid-March" and "up to six weeks" are a carrier's operational guidance, not a Chinese official measurement, and are labelled as such wherever used.
• Any published tariff of rush or expedited production surcharges. Searched; only vendor marketing pages were found. No factory, trade body or index publishes a schedule. Treat any such number you are quoted as unsourced.
• An air-against-ocean cost comparison for backpacks specifically. Air is published and dated in dollars per kilogramme. The ocean equivalent requires a chargeable-weight assumption for a container of backpacks that no primary source publishes, so the two are printed side by side and no ratio is computed.
• A named back-to-school arrival window from any customs or port authority. The June–August arrival concentration is our calculation from the monthly series, not a statement any agency publishes. The series is printed in full above so you can check the arithmetic.
• Drewry's own historical weekly series. Only the current week is published free; the history is a paid product. The 2026 values here come from verbatim republications plus one direct Drewry page, and they agree at the overlap, but no single authoritative Drewry-hosted time series was obtained.
• Port of Long Beach monthly statistics returned HTTP 403 on every attempt, so the US port evidence rests on Los Angeles alone. Port of Rotterdam's performance dashboard served placeholder values on 11 August 2026 rather than the monthly figures it names.
• Absolute published transit times for Maersk and Hapag-Lloyd transpacific services. Both publish relative claims only. Any single "Asia to North America transit time" is therefore an average of things that are not comparable.
• A current EU-wide customs inspection rate. The only sourced figure is the European Court of Auditors' range of under 1% to over 60% by member state, from 2019 data published in 2021. No newer Commission figure was retrieved.
• Our own operating figures. Everything in the "Ordering with Backper" table is self-reported and unaudited. Each line can be evidenced on request.

Sources

Import and trade series. Eurostat Comext dataset DS-045409, "EU trade since 1988 by HS2-4-6 and CN8", EU27 (2020 definition) extra-EU imports of CN 4202 92 91, indicator VALUE_IN_EUROS, flow 1, January 2022 to May 2026: the reproducible query. HMRC Overseas Trade Statistics via the uktradeinfo OData API, endpoint OTS, CN 4202 92 91, total trade. Commodity definitions from HMRC's commodity endpoint. Port of Los Angeles container statistics, loaded import containers by month.

Retail selling season. US Census Bureau, Monthly Retail Trade and Food Services, not seasonally adjusted, via the Federal Reserve Bank of St Louis republication of series MRTSSM448USN (clothing and clothing accessories stores, NAICS 448) and MRTSSM45111USN (sporting goods stores, NAICS 45111), 2022–2025.

Freight rates and surcharges. Drewry World Container Index weekly assessments, 2026, via AJOT (15 January), AJOT (11 June), Shipping Telegraph (20 February), Shipping Telegraph (10 July), Shipping Telegraph (7 August) and Drewry's own page (21 May). The 12 February 2026 assessment of USD 1,933 and the alternative 19 February dating of the week 08 assessment via Daily Cargo News (12 February) and Daily Cargo News (19 February). Golden Week 2025 commentary via AJOT, 11 September 2025. Freightos weekly freight updates: 9 June 2026, 29 July 2026 and 7 August 2026.

Ocean services and reliability. CMA CGM FAL 8 service document, marked April 2026. ONE Asia–North Europe FE1 and FE3 advisory, announced 28 May 2026. Maersk East-West Network services overview, February 2026. Sea-Intelligence press release 400, June 2026 data, published 27 July 2026.

Factory calendar. State Council Decree No. 795, published 12 November 2024, in force 1 January 2025. State Council notice Guo Ban Fa Ming Dian [2025] No. 7 of 4 November 2025, official English rendering. Hong Kong Observatory Gregorian–Lunar Calendar conversion tables for 2027 and 2028. Maersk, Chinese New Year 2026 supply chain guide, published 4 November 2025, and the blank-sailing advisory of 23 December 2025.

Ports, dwell and delay. Port Optimizer Control Tower import dwell report, snapshot 10 August 2026. Pacific Merchant Shipping Association, San Pedro Bay dwell times for June 2026. Port of Felixstowe Rates and Charges, effective 1 April 2026. US DOT Bureau of Transportation Statistics, Port Performance Freight Statistics 2026 Annual Report. European Court of Auditors Special Report 04/2021.

Manufacturing terms. Quanzhou Osgoodway Co., Ltd, published 13 May 2026. Quanzhou Huafang Bags Co., Ltd, published 25 September 2023.

Tariff. UK Department for Business and Trade Data API, dataset uk-tariff-2021-01-01, table measures-as-defined, version v4.0.1576 published 3 August 2026, for 4202 92 91 00 and 4202 92 11 00; legal base S.I. 2020/1430. EU nomenclature instrument: Commission Implementing Regulation (EU) 2025/1926, applying from 1 January 2026 — rate line not verified.

Order variance. Lee, H. L., Padmanabhan, V. and Whang, S., "The Bullwhip Effect in Supply Chains", Sloan Management Review, Vol. 38 No. 3, Spring 1997, pp. 93–102. Costantino, F., Di Gravio, G., Shaban, A. and Tronci, M., "Exploring the Bullwhip Effect and Inventory Stability in a Seasonal Supply Chain", International Journal of Engineering Business Management, Vol. 5, 2013. McGoldrick, P.J. and Betts, E.J., "Seasonal Markdown Strategies of Apparel Retailers", Proceedings of the 1997 Academy of Marketing Science Annual Conference — abstract only, no extractable figure.

Rewritten 11 August 2026. This page previously asserted a share of annual revenue, a demand multiple, an ocean transit time, an air transit time, a nearshore road transit time and a production lead time, none of which could be traced to a source, and two of which the customs data contradicts. It told buyers stock must be off the water by early July, when July is the measured peak arrival month. It implied a markdown cost for missing a season, when no usable published statistic exists. And it ended with a scarcity claim we could not evidence. Corrections have been made in the open rather than quietly deleted. Freight rates, carrier schedules, port tariffs and the Chinese holiday calendar all move — re-check anything here against the current source before it goes into a purchase order.