Separate the rates in your records. The order rate sets the commercial expectation and the selling price maths. The customs conversion, done by CBP at the rate for the relevant date, sets the duty base in dollars. The payment rate is what the goods actually cost in cash. A serious cost file holds all three and can show the difference between them as its own line, currency variance, instead of letting it smear across product margins.
That separation changes conversations. A product whose margin fell because the payment rate moved has a treasury story. A product whose margin fell because the order price rose has a sourcing story. Merged, both look like the product got worse, and the fix applied is usually the wrong one.
Decide where your standard cost takes its rate: order date is the common choice for pricing stability, with variance reported against actuals. Then state the currency terms on the purchase order itself: price currency, what happens if the rate moves beyond an agreed band before shipment, and who carries that move. Suppliers have their own exposure in the mirror image, and a clause negotiated calmly at order time beats a dispute conducted over a container already on water.
For long lead time orders, some buyers fix the rate with the supplier for a defined window in exchange for a slightly higher base price. That is buying certainty, and it can be the right buy. Price it as certainty, in the open, rather than accepting a padded quote whose padding nobody can see.
A forward contract that matches a known payment, amount, currency, and approximate date is the plain instrument for this exposure, and for most importers it is enough. The rule that keeps hedging honest is matching: the hedge exists because the payable exists, in the same size, and it settles when the payable settles. Anything larger, longer, or cleverer than the underlying orders has stopped being insurance and started being a position.
Small importers below the scale where forwards are practical still have levers: invoicing currency choice, payment timing within agreed terms, and holding a currency balance from receipts where the business earns in that currency naturally. None removes the swing. Each shrinks the amount of swing bought at the worst moment, at a retail bank spread.
Close the loop in the monthly accounts: standard landed cost per shipment, actual cash out per shipment, variance split into price, freight, duty, and currency. A variance report that names currency honestly stops it being used as the explanation for everything else, which is its fate in most margin meetings.
Review the convention yearly, not monthly. Changing the rate fixing point whenever it flatters the current quarter teaches the whole business that cost is negotiable. The convention's job is comparability. Accuracy to cash is the variance report's job, and the two should not be swapped.
Keep reading: Reverse landed cost: what a returned import really costs to bring back, Dimensional weight: when the box is light and the freight bill is not.
Pick one convention, commonly the order date rate for standard cost, and hold customs conversion and the actual payment rate as separate recorded figures. The variance between standard and cash is then visible and attributable instead of hidden in product margin.
Customs converts foreign currency values using official rates for the relevant date under its own rules, not your bank's payment rate. Duty can therefore be assessed on a dollar value that differs from what you ultimately pay the supplier.
If forwards are impractical at your size, reduce exposure structurally: choose invoicing currency deliberately, time payments within terms, and avoid concentrating many large payments on one date. Match any hedge you do use to real payables in amount and timing.
Report variance in named parts: price, freight, duty, and currency, per shipment, monthly. Once currency has its own honest line, the other lines have to explain themselves, and margin meetings get shorter.