Every operator sourcing from 1688 has felt this: you quote a product at 45 CNY per unit, run your numbers, place the order, and three weeks later when you reconcile the books, the margin is thinner than you planned and you cannot pinpoint why. Nine times out of ten, the answer is the CNY/VND exchange rate moved between the day you quoted the price and the day you actually paid for it. This post is not about explaining that mechanism (we already covered that in how CNY/VND rate swings hit your 1688 landed cost). This one gives you an operating process: three stages you run on every order, starting with the next one you place.
Why CNY/VND rate risk is the quietest killer of your margin
The gap that matters is the one between the moment you lock a price on 1688 and the moment you actually settle payment, whether that is through a payment agent, an Alipay top-up, or a forwarder invoicing you in VND. That gap is rarely same-day. Between negotiating with a supplier, waiting for a sample check, arranging QC at the warehouse, and batching payment with your agent, five to fifteen days is common. The rate does not sit still during that window.
Take a concrete order: 50,000 CNY worth of goods, quoted when the rate sat at 3,650 VND per CNY. Your landed cost at quote time is roughly 182,500,000 VND (about 7,220 USD at a rate near 25,270 VND/USD). If the rate drifts to 3,720 by the time you pay, that same 50,000 CNY now costs 186,000,000 VND, a jump of 3,500,000 VND (about 138 USD) on a single order. That is a 1.9% swing on cost. If your typical margin on that SKU sits around 18-22%, you just lost roughly a tenth of your profit on that batch, without changing a single input on the product side.
Most shop owners only catch this at month-end reconciliation, when the accounting spreadsheet shows COGS higher than expected across several orders at once. By then the damage is already booked. You cannot renegotiate with a supplier after the fact, and you usually cannot retroactively raise your Shopee or TikTok Shop price on units already sold. The loss is locked in.
This is the part most rate-risk content skips. Knowing that CNY/VND moves and hurts your margin is step one. What you actually need is a repeatable process that catches the drift before it becomes a loss, not a spreadsheet that tells you about it after the quarter is closed.
The three-stage framework: monitor, lock, buffer
Map this against the actual life of an import order. Stage one, monitoring, runs from the moment you start sourcing a product until you place a confirmed order. Stage two, locking, runs from order confirmation through final payment. Stage three, buffering, is the standing financial cushion that absorbs whatever slips through the first two stages, plus the post-mortem step where you adjust pricing if the rate has moved structurally, not just day to day.
Why split these instead of just "watching the rate more carefully"? Because each stage needs a different owner, a different tool, and a different decision. Monitoring is about information gathering, low effort, done constantly. Locking is a negotiation and timing decision made once per order. Buffering is a financial policy decided in advance, not improvised mid-crisis when the rate has already moved against you. Collapsing all three into "keep an eye on the rate" is exactly why most shops end up reacting instead of controlling the outcome.
On ownership: the shop owner sets the threshold and buffer policy (stage one thresholds, stage three buffer percentage). Whoever handles order placement and supplier communication, often the shop owner directly at small scale, owns stage two, negotiating and timing payments. Your bookkeeper or the person who reconciles COGS owns tracking actual versus planned rates and flagging when stage three buffers get triggered.
Shops running all three stages consistently report margin variance from FX alone dropping to under 0.5% per order, versus swings of 1.5-3% for shops that just glance at the rate without a structured response. That gap compounds fast if you are running 15-20 orders a month.
Stage 1: monitoring the CNY/VND rate correctly
Do not rely on a single source. There are three rates that matter and they are not the same number. The State Bank listed rate is the official reference but rarely what you actually pay. Black market or free-market rates (cho den) move faster and often more accurately reflect near-term direction, but they carry their own spread. The rate your payment agent or Alipay top-up service actually charges you is the one that hits your P&L, and it typically sits somewhere between the two, with a service margin baked in that can range from 0.5% to 2% depending on the provider.
Track all three, but budget against the agent's actual rate. If you only watch the official rate, you will consistently underestimate your real cost.
Frequency should scale with order size. For orders under 20 million VND (roughly 790 USD), a weekly glance at the trend is enough, the absolute VND exposure from a 1-2% swing is small in cash terms. For orders in the hundreds of millions of VND, check daily, and check the specific agent rate before you authorize payment, not just the general market rate.
Set a concrete alert threshold instead of a vague "watch it closely." A reasonable default: flag anything that moves more than 1.5% from the rate you used when you first quoted the order. At 1.5% on a 100,000 CNY order, that is roughly 5,475,000 VND (about 217 USD) of exposure, enough to matter, not so tight that you get false alarms on normal daily noise.
Logging does not need software. A simple spreadsheet with four columns, date, agent rate, official rate, and order reference if applicable, gives you the historical data to calculate real margin later instead of guessing. This same log becomes the input for the buffer calculation in stage three. If you already track full landed cost for 1688 orders, add rate-on-quote-date and rate-on-payment-date as two extra fields in that same sheet.
Stage 2: locking the CNY rate when you place the order
Within a standard 1688 ordering process, there are three natural checkpoints where you can lock a rate: deposit payment, final balance payment, and goods received at the China-side warehouse before forwarding. Each one is a different commitment level.
Locking at deposit is the most aggressive move, you fix the rate on maybe 30% of the order value while the remaining 70% still floats. Locking at final payment means you carry full exposure until the order is essentially done on the supplier side. Locking at warehouse receipt only makes sense if your forwarder offers a combined payment-and-freight rate that bundles FX, which some larger agents do for repeat clients.
Talk to your payment agent or forwarder directly about holding a rate for a fixed window, commonly 3-7 days. Agents handling volume for repeat clients will often agree to this informally, especially if you are paying in batches through them regularly. It costs you nothing to ask, and a locked window removes the guessing entirely for that order.
Pay in full upfront to lock early when you have high confidence in the order (repeat SKU, trusted supplier, stable demand) and cash flow allows it. Pay in installments instead when the order is new, unproven, or large enough that tying up capital for three extra weeks creates its own risk, in that case managing cash flow on bulk 1688 orders matters more than shaving a fraction of a percent off FX exposure.
The two mistakes that show up most often: locking a rate before the order is actually confirmed with the supplier (you eat the FX cost of a deal that might fall through), and the opposite, letting payment drift for two or three weeks after price negotiation because you are waiting on other line items to bundle into one payment run. Both cost real money. Set a rule: lock within 48 hours of supplier confirmation, no exceptions for convenience.
Stage 3: buffering your margin against rate risk
Your buffer should come from actual data, not a round number picked because it feels safe. Pull twelve months of CNY/VND movement (your stage one log gives you this once it has run a few months; before that, use published historical data). Calculate the standard monthly swing. Over the past two years, CNY/VND has typically moved 1-3% within any given month, with occasional spikes past 4% during periods of broader VND pressure. A buffer set at 2% covers most normal months; shops carrying larger average order values or longer supplier lead times should run closer to 3%.
The adjustment formula is simple: when the rate you are about to pay exceeds your stage-one alert threshold (the 1.5% mentioned earlier), recalculate landed cost using the actual rate, not the quoted one, and flag the SKU for a pricing review if the new landed cost pushes margin below your minimum acceptable threshold, commonly 15% for most SEA cross-border operators.
Keep the buffer as a separate line in your cash flow, not folded into general import capital. If it is mixed in, it gets spent on the next order instead of sitting ready to absorb a bad rate month. A dedicated FX buffer account, even a simple separate bank sub-account, holding 2-3% of your average monthly import volume, keeps this real instead of theoretical.
On when to actually move your TikTok Shop or Shopee price: only after the rate move looks structural, not a one-week blip. If VND has weakened against CNY for three consecutive weeks and your buffer is getting drawn down without recovering, that is a signal to adjust. Move price in small increments, 3-5% at a time, and time it with a normal restock cycle rather than a sudden mid-cycle jump, which reads as erratic to repeat buyers and tanks conversion on marketplace algorithms that weight price stability.
Case study: running the framework on a real order
Order: 80,000 CNY of kitchenware, sourced from a Guangzhou supplier. Quote date rate: 3,660. Deposit paid three days later at 3,668 (within threshold, no action). Final balance paid twelve days after quote, rate had moved to 3,745, a 2.3% swing, past the 1.5% alert.
Without the framework: shop pays the full 80,000 CNY at 3,745 with no adjustment, landed cost 299,600,000 VND against a planned 292,800,000 VND, a 6,800,000 VND (about 269 USD) overrun absorbed silently into thinner margin. Nobody notices until month-end.
With the framework: stage one flags the 2.3% move against the deposit-day rate. Stage two had already locked 30% of the order at deposit (3,668), limiting exposure to the remaining 70%. Stage three buffer, funded at 2% of monthly volume, absorbs 4,700,000 VND of the overrun without touching the core margin on that SKU. Net effect on final margin: roughly 0.6 percentage points lost instead of 1.9 points lost without the framework, on this single order.
The single step that mattered most here was the deposit-stage lock in stage two. It capped exposure on 30% of the order before the rate moved at all. Monitoring told the shop something was happening; locking is what actually reduced the dollar amount at risk.
For shops just starting out with small capital, do not over-engineer this. Skip the separate buffer bank account at first if your order sizes are under 20 million VND per run, the absolute exposure is small enough that a simple spreadsheet flag and a mental 2% price cushion covers you. Scale the formal buffer fund once your monthly import volume crosses roughly 200 million VND (about 7,900 USD).
Tools that help with tracking and locking rates on 1688
Real-time price and rate tracking matters most during stage one, when you are sourcing and comparing suppliers on 1688 and want to know at a glance whether the CNY/VND rate you are quoting against is still current, rather than checking a browser tab separately every time.
Once an order moves into your actual import pipeline, the value shifts to payment and rate management: knowing which rate you locked, when, and against which order, without digging through chat logs with your agent.
If you are running a handful of orders a month, a manual spreadsheet covers all three stages fine, the framework does not require software to work. What software buys you is speed and fewer missed thresholds once order volume climbs past what one person can track by memory.
We are building toward this at Ordinex. Scout, our sourcing tool, is in private beta and aimed at exactly the stage-one problem: seeing price and rate context together while you are sourcing on 1688, instead of switching tabs. Orders, our import workflow tool, is also in private beta and covers stage two and three, tracking payment timing against locked rates per order. Neither tool replaces the discipline of running the framework yourself. They just make it faster to run once you have more orders than a spreadsheet can comfortably handle.
Frequently asked questions
How much does CNY/VND have to move before it meaningfully hits margin? Anything past 1-1.5% is worth acting on for most shops. Below that, it is usually noise within your existing cost buffer. Above 2%, on a typical 15-20% margin SKU, you are looking at losing a meaningful fraction of that order's profit if left unaddressed.
Should I lock the rate at deposit or wait for full payment? Lock at least the deposit portion once the supplier has confirmed the order. Waiting for full payment to lock anything means you carry 100% exposure for the entire order cycle, which is the riskiest option, not the safest one.
Is a forward rate lock service worth it for small orders? Generally no. Forward products from banks or larger FX desks usually have minimum ticket sizes that do not make sense below roughly 500 million VND (about 19,800 USD) per transaction. For smaller orders, negotiating a short hold window with your payment agent achieves most of the same protection for free.
Where should I check the CNY/VND rate to get the most accurate number? Check the rate your payment agent or Alipay top-up provider actually quotes you, not the official interbank rate. Cross-reference against a live market tracker for direction, but budget against the agent number since that is what actually lands on your invoice.
What buffer percentage is reasonable for FX risk? 2% of order value as a standing buffer covers most normal months based on historical CNY/VND volatility. Shops with longer supplier lead times or larger average order sizes should run closer to 3%.
If you are ready to put this framework into practice, Scout and Orders are both open for private beta sign-up at ordinex.cc. Start logging your rate on the next order you quote, even before any tool. That single habit is what makes the rest of this framework possible.