Cross-border e-commerce stock planning is a forecast bounded by customs clocks
Cross-border e-commerce stock planning in China is the discipline of deciding how many units of each SKU to hold in a bonded warehouse, at which port, and when to replenish, used to keep a 1210 storefront in stock without holding goods past the point the customs rules allow. It looks like ordinary inventory planning and is not: an ordinary warehouse lets you sit on slow stock until the market turns, while a bonded position has three legal clocks running against it from the day the goods clear into the zone.
The commercial logic is different too. In bond, the import tax is not yet paid, so a unit's carrying cost is warehouse space plus the cost of the goods — not the duty and VAT that a general-trade import would have settled at the border. That is precisely why bonded stock is the default first position for a foreign brand testing China, and why underestimating the clocks is the most expensive mistake in the model.
Bonded stock defers tax rather than removing it
A bonded retail import position is inventory held inside a customs-supervised zone on which duty and import taxes have not yet been assessed, used to serve consumer orders that are taxed at the moment of release rather than at the moment of entry. China's cross-border retail import tax policy sets the parameters: RMB 5,000 per order and RMB 26,000 per person per year, zero tariff inside those limits, and import VAT plus consumption tax at 70 percent of the statutory rate (Notice 49 of 2018, Ministry of Finance).
Two consequences follow for the replenishment plan. First, the cash-flow argument for bonding is real but bounded — the benefit is a deferral, so the gain scales with how long the stock sits and how much tax it would otherwise have settled. Second, the price ceiling per order is a hard cap, not a guideline: a cart that crosses RMB 5,000 falls out of the retail import channel, so high-ticket SKUs need a different route rather than a bigger bonded position.
The positive list is the boundary of every purchase order
The cross-border e-commerce retail import positive list is the schedule of tariff lines eligible for the 1210 and 9610 channels, used to screen a product range before any replenishment order is placed. The 2022 edition carried 1,476 tariff lines and took effect on 1 March 2022, and a line that is absent cannot be sold through the bonded route however good the demand looks in a forecast.
Inside the list, the regime is unusually fast. Notice 486 of 2018 waives first-import licensing, registration and record-filing for goods bought through cross-border retail channels, provided the seller publishes a risk disclosure on the order page and gives consumers access to a Chinese electronic label (Notice 486, State Council). For a replenishment planner that means a new SKU can go live in weeks, with no registration file to wait on — so the binding constraint on assortment is the list, not a regulator's queue.
Three customs clocks bound the position
A bonded stock clock is a customs-set deadline that starts when goods enter or leave the zone, used to decide the maximum time a unit may sit in the position before it must be sold, returned or destroyed. All three clocks run simultaneously, and the shortest one governs.
| Clock | Rule | What it forces into the plan |
|---|---|---|
| Storage | One year for bonded warehouse goods, extendable once by up to one year with customs approval (GACC Order 105, as amended) | No position may be planned on a sell-through horizon longer than 12 months |
| Destruction | 60 days from the customs-signed application, one extension of up to 30 days, video evidence kept 3 years (GACC Announcement 79 of 2025) | Dead stock must be adjudicated while it still has resale value, not after |
| Returns | Application within 30 days, arrival within 45 days, goods fit for resale (GACC Announcement 45 of 2020) | Returned units re-enter available stock and must be netted off the next order |
The storage rule comes from the bonded warehouse regulation, which also obliges the operator to keep a true monthly record of goods received, released and remaining, and to report it to customs, which may inspect at any time (GACC Order No. 105, consolidated text). In practice the monthly stock statement is the document your replenishment plan has to agree with — a plan that exists only in a spreadsheet and not in the bonded ledger is not a plan.
Sizing the first position backwards from the quota
Position sizing is the calculation of how many units of a given SKU to place in bond at launch, used to cover a target reorder cycle without breaching the storage clock or the per-customer quota. Work it in four steps.
- Divide RMB 26,000 by your landed retail price to get the maximum units one customer can buy in a year; a RMB 380 product caps at 68 units, a RMB 1,200 product at 21.
- Multiply by a realistic repeat rate to get expected annual units per buying customer, then by the number of buying customers the launch media plan can plausibly reach.
- Divide by 365 and multiply by your cover days — 45 to 60 days for a launch, longer only where a campaign is scheduled.
- Add the campaign uplift, subtract stock already in bond and units in transit that will land inside the cover window.
The step most brands skip is the first. Because the quota is per person, a bonded position cannot be sold deeply into a small group of enthusiasts; it has to be sold broadly. That pushes the assortment decision towards more buyers per SKU rather than more units per buyer, and it argues for a narrow launch range with a defensible hero product rather than a full catalogue shipped into bond.
Flat shopper numbers make replenishment a share fight
Cross-border import demand in China is the volume of consumer goods bought through the 1210 and 9610 channels, used as the ceiling against which any replenishment plan must be tested. China's cross-border e-commerce imports were worth RMB 570.2 billion in 2025, up 2.7 percent, and the mix was entirely consumer goods: beauty, fragrance and personal care at 27.2 percent, food and fresh produce at 23.9 percent (GACC, 2025 cross-border e-commerce statistics).
The demand pool is not expanding fast enough to bail out a bad buy. Online shoppers reached 976 million in June 2025, up 0.1 percent in six months (CNNIC, 56th Statistical Report). When the shopper count is flat, every incremental unit sold in bond has been taken from another product on the same shelf — so replenishment should concentrate depth on SKUs with proven conversion and treat breadth as an experiment with a fixed budget, not as a hedge.
Lead time and campaign peaks set the reorder point
Replenishment lead time is the elapsed time from purchase order to sellable bonded stock, used to place the reorder point far enough ahead that cover never breaks. Build the number from production time, ocean transit, port handling, bonded putaway and the platform's listing checks, then add the buffer that the season requires.
Seasonality is where bonded planning differs most from general trade. Stock for a June or November campaign has to be in the zone weeks before the peak, but ordering early burns the storage clock and locks capital into a forecast. The resolution is to split the buy: a core position sized on baseline demand that lands early, and a campaign top-up that lands close to the peak and is priced to clear. Brands that order one large container for the peak usually end up with either a stockout in week one or aged stock in January — see our China campaign calendar for how the peaks fall, and customs clearance timings for the port-side variable.
Five metrics for the monthly replenishment review
A bonded replenishment review is the monthly meeting at which stock cover, sell-through and age are compared against the plan, used to issue the next purchase order and to adjudicate dead stock. Five numbers carry the decision.
- Days of cover by SKU and by port — one number per position, not one national average.
- Sell-through at day 90 on each launch SKU, measured on units released against units landed.
- Aged stock beyond 180 days, valued at cost, because this is the population heading for destruction.
- In-stock rate on the top 20 SKUs, which is the only service metric a customer experiences.
- Write-off at destruction, tracked as a cost line rather than absorbed as shrinkage.
How GOODSINFINITE runs the bonded position
GOODSINFINITE acts as importer of record and bonded fulfilment partner across five locations — Tianjin, Shanghai, Ningbo, Guangzhou and Qingdao — which puts the replenishment decision where the customs filings are: the monthly stock statement, the release-by-release tax settlement, the return filings and, where it comes to it, the destruction application. Brands keep title to their stock; we run the position with them on a monthly cycle so that the reorder point is set on measured cover rather than on a container's worth of optimism. For the wider sequence, start with our China market entry guide and the cross-border operations glossary.
FAQ
How much stock should a foreign brand hold in a Chinese bonded warehouse? Size from demand and from the quota ceiling. Because orders cap at RMB 5,000 and each customer at RMB 26,000 a year, one buyer can absorb only a limited number of units, so a launch position should cover 45 to 60 days on the top 20 percent of SKUs and a small sample on the tail, held at the port whose catchment matches the launch region.
How long can 1210 bonded stock stay in a bonded warehouse? One year as a rule, with a single extension of up to one year where customs agrees there are justified reasons. Treat twelve months as the planning horizon: a unit that cannot sell through inside it must be exported, transferred to another special supervision area, converted to general trade with duty paid, or destroyed.
What happens to bonded stock that does not sell? It is destroyed under customs supervision. Announcement 79 of 2025 requires destruction within 60 days of the customs-signed application form, allows one extension of no more than 30 days, requires video evidence kept at least three years, and files the disposal under code 0200 — or 0844 with tax paid if the destruction generates revenue.
Does bonded stock reduce the import tax a brand pays? It defers and discounts it. Inside the RMB 5,000 and RMB 26,000 limits, tariff is zero and import VAT plus consumption tax are levied at 70 percent of the statutory rate, an effective 9.1 percent on a 13 percent VAT product and near 23.1 percent on high-end cosmetics. Nothing is payable until release.
Can any product be replenished into bonded stock? Only tariff lines on the positive list — 1,476 lines in the 2022 edition effective 1 March 2022. Inside that boundary, Notice 486 of 2018 waives first-import licensing and registration, so SKUs can be added in weeks, provided the listing carries the risk disclosure and Chinese electronic label.
Sources
- GACC — Announcement No. 79 of 2025 on destruction and return of cross-border e-commerce bonded stock: 60-day destruction, 30-day extension, 3-year video retention
- GACC — Order No. 105, Rules on Bonded Warehouses and Goods Stored Therein: one-year storage, monthly stock reporting
- MOFCOM et al. — Notice 49 of 2018 on the tax policy for cross-border e-commerce retail imports: RMB 5,000 per order, RMB 26,000 per person, 70 percent of statutory VAT
- State Council — Notice 486 of 2018 on improving supervision of cross-border e-commerce retail imports: no first-import licensing, risk disclosure, electronic label
- GACC — 2025 cross-border e-commerce statistics: imports of RMB 570.2 billion, beauty 27.2 percent, food and fresh produce 23.9 percent
- CNNIC — 56th Statistical Report on China's Internet Development: 976 million online shoppers, 86.9 percent penetration