Best Time to Exchange Currency: What Actually Sets What You Pay

The short answer

Ask a broker, a treasury desk or an economist when to convert and you get the same reply: it cannot be forecast. Freely traded rates move on interest-rate expectations, inflation data, politics and flows, and no participant's information is enough to pick a day. That is not a hedge — it is the answer to this search.

The question still matters, because timing is not the expensive part. For a trip, a tuition payment or a supplier invoice the amount is fixed and the conversion happens once, so the gap between this week's rate and next week's is small beside the gap between a good channel and a bad one. Where and how you convert is a decision you can make; when is not.

So replace “when is the best time” with questions you can answer from public information: what is the mid-market rate, what are you actually being offered, and what is the gap as a percentage of the money moving. Once you can compute that gap, the calendar stops mattering.

• Do not forecast the rate. The channel you use costs more than the day you use it.
• Real timing facts: card networks post on business days, weekend quotes can carry a wider margin, and airport and hotel counters price the same currency worse.
• The one timing choice you really face is a merchant offering to convert on the spot. Decline it.

The three numbers behind every quote

Every quote you will be offered is a version of the same thing: a benchmark rate with margin built in, plus sometimes a fee. Providers present those parts differently on purpose, so step one is knowing what each one is, and step two is the calculation that makes them comparable.

• Spread: buy dearer than mid, sell cheaper — a round trip crosses it twice.
• Markup or fee: the part that hides, depending on who is quoting.
• Your number: total cost as a percentage of the amount moved.

The mid-market rate

The mid-market rate is the theoretical exchange rate between two currencies with no dealer margin — the figure you see when you look a pair up, and the one approximated by very large interbank trades between institutions dealing in each other's currency rather than with the public.

Treat it as a reference, not a price. No consumer transaction settles at it: there is always something on top, whether or not anyone calls it a fee. And it is not one official number — vendors publish slightly different figures for slightly different moments, and a pair can be quoted in either direction. Use one reference for both quotes and the comparison holds.

The spread

The spread is the gap between the rate at which a dealer buys a currency and the rate at which it sells, with the mid-market somewhere between them. You cross it every time you convert, and you cross it twice if you change money into a currency and later change the leftovers back. A rate that looks friendly on one side usually hides the cost on the other, so compare the round trip.

The markup, and the fee that is not a fee

A markup is margin expressed as a percentage of the rate or as a flat charge. Some providers quote an attractive rate and add a separate commission or transfer fee; others fold everything into the rate and advertise no fee. Both are the same money, and the second presentation is why “no commission” on a bureau sign tells you nothing about price.

The one calculation that settles it

Take the amount you were actually charged, in your own currency, for a fixed amount of foreign currency. Divide it by what that same foreign amount would have cost at the mid-market rate. Subtract 1. That is your true cost as a percentage — (what you were charged ÷ mid-market) − 1 — and it works on a card statement, a bureau receipt, a transfer confirmation or an app quote.

Two details decide the answer. Both figures must describe the same foreign amount: the cost of the money you ended up with, not what you started with. And both must be in your own currency, so invert the mid-market rate if it is quoted the other way round.

With rates instead of totals, divide the rate you were sold at by the mid-market rate for the same direction and subtract 1 — safe only when both are quoted as units of your currency per one foreign unit. For a trip plan, multiply your budget by that percentage for the conversion's cost in money.

Paying by card, exchanging cash, or sending a transfer

The three ways most people exchange money price completely different things. Once you see which costs are unavoidable in each and which are yours to choose, the “best time” framing collapses into a channel decision.

Paying with a card, or using an ATM abroad

A card purchase abroad is a conversion with extra parties: the merchant's acquirer, the card network, your issuer and sometimes a processing agent each touch it, and any of them can add margin. On weekdays you typically see a rate close to the mid-market benchmark plus, depending on the product, an issuer charge — some levy one and some do not, and only your own card's terms bind you.

What you control is the on-the-spot choice. When a terminal, an ATM screen or a checkout offers to charge you in your home currency instead of the local one, that offer is itself a conversion at the local side's rate, and the margin inside it is why the offer gets made. Declining it leaves the conversion with your own issuer, whose basis you can read in advance.

What you do not control is the date. Card networks post and settle on business days, so a weekend or holiday purchase converts and settles on a later weekday, and which day's rate applies depends on your issuer's process. A large payment you can move yourself — a car rental, school fees — is worth paying on a normal business day. ATMs add a charge purchases do not carry, so several small withdrawals lose to one larger one.

Cash at an airport, a hotel desk or a bureau

A physical counter holds inventory, pays rent and takes risk, and prices all three into the rate rather than into a fee. Airport and hotel counters are consistently dearer than comparable high-street bureaux for structural reasons: a captive, time-pressured customer base, expensive floor space, and stock that has to sit on site against unpredictable demand. The direction is predictable even when the size is not.

Two consequences. If you need notes on landing, buy only what gets you to a city and do the rest somewhere with competition around it. And watch the second leg: leftover notes have to be changed back, often at a worse rate than you bought at, and counters may refuse worn notes or older series. Always ask the all-in question: how much of my currency for how many units of that one.

Sending money as a transfer

For anything that has to arrive as money in another country — rent, invoices, family support, a deposit — the margin inside the rate and any flat charge trade off against each other. A tight rate with a fixed fee can be dearer on a small amount than a looser rate with none, and cheaper on a large one, so the question is never “who is cheapest” but “who is cheapest on my amount”.

Check what arrives, not what you sent. Correspondent and intermediary charges can reduce the credited sum, and a quoted arrival date assumes a business day in both countries. If the payment must land on a date, the rate is secondary to certainty.

Timing facts that are structural, not predictions

A handful of calendar effects are real because they come from how providers operate, not from where the rate is going. None tells you direction; they tell you when a quote is likely dearer than a weekday one, and why.

The mechanism is nearly always the same. When the interbank market is shut, a provider quoting you holds currency risk it cannot hedge until reopening, and prices that risk into the rate. Some card and app products make this explicit with a weekend or out-of-hours margin during published windows; others leave it in the spread. Which one you use is a question for its own terms page.

Everything else follows from closures and deadlines: shorter hours around public holidays, a bureau shut, a transfer sitting unprocessed until the next business day in whichever country is on holiday. A fixed arrival date is governed by cut-off times and two banking calendars — the only part of this you can win, and worth paying a worse rate for.

• Card networks post and settle on business days, so weekend purchases convert on a later weekday.
• Weekend and out-of-hours quotes can carry a wider margin, because the provider cannot hedge while the market is shut.
• Some card and app products state a weekend or out-of-hours surcharge explicitly — check yours.
• Fixed-date payments: the cut-off time and the calendar beat the rate.

Does the best time differ by country? UK, Malaysia, Australia, Singapore

Much of this search is regional, and the honest regional answer is that three things vary by country and none can be settled for a general audience: what a local counter can stock, whether you get a better deal buying the destination currency before you leave or after you arrive, and whether that currency is restricted or quoted with a wide margin. The UK, Malaysia, Australia and Singapore each have their own product line-ups and rules, and those change.

For what genuinely is national, go to a primary source rather than a forum: the destination's central bank or financial regulator publishes what is legal tender, what is restricted or must move through licensed channels, and what disclosure rules quotes fall under. Your bank's tariff covers the other half. If a currency is restricted or thinly traded, no choice of day fixes that.

• What a bureau can stock: major currencies are everywhere; less common notes may need ordering, or only exist at the destination.
• Where the deal is better: varies by pair and place, so price the same amount at home and locally on the same day.
• Where to check: the destination's regulator or central bank, plus your bank's own tariff page.

Cards and apps: what “best time on Revolut” is really asking

Queries about specific card and app products are one question in different words: at what point does this product stop being cheap. The general pattern is structural, not a forecast. An app-style product typically quotes close to the mid-market rate during weekday market hours; it may add a margin at weekends and out of hours, when the market is shut and it carries the risk itself; and free plans commonly cap how much you can exchange or withdraw before charges start.

This page deliberately gives you no numbers for any company. Terms, allowances and surcharge windows move, and the version that applies to you is the fee schedule attached to your own account. Read four things there: which rate basis is used, when a margin applies instead, what the free allowance for exchange and ATM withdrawal is and how it resets, and what happens once you exceed it. Two questions people skip: what the product does when you pay in a currency it does not hold, and what it does at weekends. Both are where a good headline rate turns into an ordinary one.

How to compare two quotes in one minute

Everything above reduces to one procedure. Fix the amount of foreign currency you want to end up with — it is the only number both quotes agree on. Price that amount once at mid-market, using a single reference for both. Then turn each quote into one percentage. Every figure below is hypothetical, chosen so you can check the arithmetic; the method is what transfers.

• Fix the foreign amount you want to end up with.
• Divide each quote's total charge by that mid-market figure.
• Subtract 1, and compare percentages instead of rates or fee lines.

A card statement after the fact

You were billed 500 EUR at a hotel and the statement shows 561 USD. At the reference rate of 1 EUR = 1.10 USD, those 500 EUR would have cost 550 USD. So 561 ÷ 550 = 1.02, and that conversion cost you 2 percent.

Same method on a rate instead of a total: if mid-market is 1.10 and your bank quoted 1.133 for the same euros, then 1.133 ÷ 1.10 = 1.03 — 3 percent — whether or not anyone described it as a fee.

Two cash counters, and the second leg

You want 1,000 EUR in notes. At 1 EUR = 1.10 USD the mid-market price is 1,100 USD. Counter A charges 1,133 USD for those 1,000 EUR: 1,133 ÷ 1,100 = 1.03, so 3 percent. Counter B advertises no commission and charges 1,155 USD: 1,155 ÷ 1,100 = 1.05, so 5 percent. The no-commission counter is dearer by 22 USD, since 1,155 − 1,133 = 22.

Now the leg nobody quotes you. Suppose you buy 1,000 EUR at 1.15, spend 700 EUR, and change the remaining 300 EUR back at the same counter at 1.05. Those 300 EUR cost 345 USD and come back as 315 USD, so 345 − 315 = 30 USD lost on currency you never used. That is the price of converting twice, and no single quote shows it.

A transfer with a fee

Suppose 1 USD should buy 15.00 units of the destination currency at mid-market and the recipient must receive 15,000 units, which is 1,000 USD at that rate. Provider one delivers 15,000 units for 1,050 USD: 1,050 ÷ 1,000 = 1.05, so 5 percent, with no stated fee. Provider two delivers the same 15,000 units for 1,020 USD plus a flat 25 USD, and 1,020 + 25 = 1,045.

Against the same 1,000 USD of mid-market value, provider two's all-in figure is 1,045 ÷ 1,000 = 1.045 — 4.5 percent — so it wins even though its rate alone was worse. Divide a flat fee by the amount sent to see why: 25 ÷ 1,000 = 0.025, which is 2.5 percent, while 25 ÷ 10,000 = 0.0025 — a quarter of one percent.

The arithmetic trap

Never subtract two rates and call the difference the cost. If mid-market is 1.25 units of your currency per foreign unit and a counter sells that unit at 1.50, the gap looks like 0.25 but the real figure is 1.50 ÷ 1.25 = 1.20, so 20 percent. Flip the direction and a same-looking gap answers differently: the counter buys your unit back at 1.00, and 1.00 ÷ 1.25 = 0.80, so you receive 20 percent less than mid. Always divide, then subtract 1.

Mistakes that cost more than the timing

The money in this decision goes to a small set of avoidable errors, and none of them involve the calendar.

• Accepting a merchant's or ATM's offer to charge you in your home currency. Paying in local currency keeps the conversion with your issuer, whose terms you can read in advance.
• Converting twice — buying notes you do not spend and changing them back — which pays the spread on both legs.
• Comparing quotes on their visible half: a tight rate with a fat fee against a loose rate with none.
• Reading a review site's fee table instead of the terms attached to your own account.
• Ignoring a fixed arrival date to chase a rate, and paying for the delay as well.

Using this converter to price a quote

The linked tool fits here, and narrowly on purpose. It is arithmetic, not a price feed: it takes an amount, a pair and your inputs and returns the two figures whose ratio is that percentage. Use it after you have quotes, not while you hunt for a day to buy.

Enter the amount and the pair for the mid-market result and the rate used. Put the rate your bank or bureau actually quoted into the manual rate override — the right move for a pair the page does not stock, or when you want their price rather than the snapshot's. Then set the bank markup percentage and the page shows the mid-market result, the bank result and the hidden fee between them.

Because every cross rate comes from one single-USD table, EUR to JPY and JPY to EUR are the same trade, with no inconsistency between directions — the property you need when neither provider quotes against your home currency. Export the CSV to line up several quotes later.

• Ten currencies: USD, EUR, GBP, BRL, IDR, JPY, CNY, AUD, CAD, CHF.
• Manual rate override for the rate you were actually quoted.
• Bank markup percentage: mid-market result, bank result, hidden fee.
• CSV export, no API calls, nothing uploaded.

What the page gives you

Inputs: an amount, From and To selects, a swap control, a manual rate override, and a bank markup percentage running 0 to 10 with a default of 1.5. Outputs: the mid-market result, the bank result, the hidden fee between them, the rate used, a warning if you combine an overridden rate with a markup, and the rendered rate-snapshot table.

Rates come from a hardcoded static mid-market snapshot dated 2026-09-15, expressed against USD, covering ten currencies: USD, EUR, GBP, BRL, IDR, JPY, CNY, AUD, CAD and CHF. The page makes no API calls at all and runs entirely in your browser, so nothing you type is sent to any server.

What this tool cannot tell you

It will not answer the question you arrived with, and saying so is part of the job. There is no live feed: the rates are a static mid-market snapshot dated 2026-09-15, and the tool's own FAQ states that they are static mid-market values which do not update in real time. Need today's rate? That FAQ sends you to your bank or a licensed FX provider.

Three limits follow. The snapshot may sit far from today's market, so use it to compare quotes on a consistent basis, never as a price. Only those ten currencies exist, so an unsupported pair has to be looked up elsewhere and typed into the manual override. And the tool accounts for the rate only: no wire charges, no intermediary deductions, no taxes, no ATM fees.

It also has nothing to say about timing: no historical series here, and no forecast of one. Everything on this page is planning information about costs and channels — not advice to buy or sell any currency, and not a view on where a rate is going.

Frequently asked questions