Currency Conversion Traps

Spread definition

An exchange spread is the distance between the rate a currency trades at and the rate applied to a transaction. It is charged inside the conversion rather than added as a line, which is why it rarely shows up on a checkout screen as a cost at all. The model used here quotes a spread as a percentage, applied to the amount converted, so the same spread costs more on a larger order without its stated size ever changing.

Two conversions can land on a single order. The first takes place where the listing currency differs from the currency the platform settles in. The second occurs when the settlement currency and the currency your account holds are not the same, and there the issuer applies a rate of its own. Each step carries its own spread, and the second is frequently invisible until a statement posts. Where all three currencies match, the spread is zero and the order costs exactly what the screen said.

Conversions between a settlement currency and an account currency carry a charge that this desk models at roughly two and a half per cent. The figure is a working parameter carrying a review date rather than a quotation from any bank, and it should be replaced with whatever your own statement shows before it is used to decide anything. What does not change with the parameter is the structure: every conversion step is a multiplicative cost, and steps multiply rather than add.

Conversion arithmetic

Conversion arithmetic is ordinary multiplication with one trap: the order of the operations is not visible on the screen. Take the amount in the listing currency, apply the first conversion where one is required, then apply the second where the account currency differs again. The result is what leaves the account, and the distance between that figure and the listed price is the cost of moving money rather than the cost of goods. Writing both steps out shows which one is worth removing.

A worked example makes the size of it concrete. On a hundred-and-eighty-dollar order converted once at the model spread, the conversion cost is roughly four and a half dollars. Converted twice, the second step charges its own percentage on an amount that has already been converted, so the combined cost rises while the order itself has not changed at all. Nothing about the goods, the freight or the duty moved; only the route the money took.

Rounding sits underneath the arithmetic and is easy to overlook. Rates are applied when the payment is captured rather than when it is authorised, so a rate shown at checkout is not the rate that posts. Amounts are rounded to the minor unit at each step, and a refund reversed on a later date crosses the spread a second time in the opposite direction. The honest estimate is therefore a range, and the width of that range comes from how many conversion steps are involved.

Conversion also feeds lines that have nothing to do with payment. Where duty is assessed on the goods plus the freight that delivered them, a weaker rate applied to a larger converted figure enlarges the base that the destination band is then applied to, so one conversion cost can be multiplied by whatever the lane charges. The effect is slight on a low-band lane and noticeable on a high-band one, which is a further reason to record converted figures rather than only the listed price.

Two currency walkthrough

Take a listing priced in dollars and an account funded in euros. Paying in dollars means the platform converts for settlement and the issuer converts again when the charge reaches the account, so two spreads apply and neither is visible at the point of payment. Paying in euros, where the option exists, removes one step but hands the conversion to the platform at its own rate, which may be better or worse than the issuer's. The comparison turns on which of the two rates sits closer to the rate the pair actually traded at.

A second case involves a listing priced in one currency and an account funded in a third. Here the arithmetic has three steps rather than two, and a further problem appears: the platform may not offer the intermediate currency at all, so the choice becomes one between a longer conversion chain and a different rail entirely. Reading the checkout screen for the currency actually being charged is the only reliable way to see which chain is in play.

Both cases finish in the same place, which is the statement line. The amount displayed at payment and the amount posted after settlement are two different numbers, and what separates the two is the whole conversion cost, with nothing hidden inside it. Recording that gap once per currency pair converts a vague suspicion about conversion into a measured figure, and it takes two lines in a note: what was shown, and what was charged.

Markup hunting

Hunting a markup begins with the rate used rather than the amount charged. Most statements and platform receipts carry a rate line or an equivalent disclosure; where one exists, compare it with the rate the pair traded at on the same day from any public source. The difference is the markup, expressed as a percentage, and that percentage is the number worth carrying into a decision. A suspicion that conversion is expensive cannot be acted on; a measured gap can.

A cheap test makes the measurement routine. Pay one small order in a currency that requires a conversion, then record both numbers on the day the statement posts. Repeat once with an order paid in the account currency. The two records now differ by the conversion cost alone, because the goods and the freight were held roughly constant, and the comparison describes your own account rather than a published band that may not apply to it.

Two habits remove most of the avoidable cost. Choose the currency at checkout instead of accepting a default, since a default usually adds a step. And keep refunds in view, because money coming back crosses the same spread in reverse, so a refund issued weeks later can return less than it took. Neither habit requires a different rail, and together they are worth more than any comparison of headline surcharge bands.

Rate snapshot

A rate snapshot is a dated note holding three figures: the rate the pair traded at, the rate applied to the transaction, and the resulting markup. One snapshot is an anecdote; a dozen across two currency pairs describe how an account is priced. Snapshots belong in the same folder as order records, because a conversion question raised in a dispute is answered by a dated pair of numbers rather than by anybody's recollection of what the rate used to be.

Snapshot discipline is mostly about dates. A markup computed from a rate read a week later is not a markup but a guess, since the pair may have moved further than the spread being measured. Read the rate on the day the charge posts, write the date beside it, and store the source of the rate alongside for the same reason. Snapshots older than the review window belong in an archive rather than in the working note.

This desk publishes parameter bands rather than live rates, and that difference shapes how the numbers should be used. A band says what order of magnitude to expect and how to structure a comparison; only your own statement says what you were charged. Readers who want the conversion carried through a complete order can enter both figures in the landed-cost waterfall, where the spread sits as one editable line among the seven it sums.