The short answer
A forex swap is the overnight financing charge or credit your broker applies when you hold a leveraged position past the daily rollover, set by the interest-rate difference between the two currencies, because holding a pair means borrowing one and lending the other. The net of that borrowing and lending is the swap that lands on your account each night, small per night but decisive over a long hold.
I want to lead with the practical split, because it decides whether swap matters to you at all. A day trader who closes before the rollover never pays a swap, and a swing or position trader who holds for weeks feels it as a real cost or credit that compounds quietly against, or into, the trade.
The honest framing is that swap is the rent on an overnight leveraged position, and like any rent it is small per night but decisive over a long hold, and ignoring it is how a flat-looking trade bleeds to death by a thousand daily debits.
The wider context is in the forex basics guide, and this page covers the overnight cost that the leverage and position-sizing pages reference throughout.
What a swap actually is
When you hold a forex position overnight, you are holding a leveraged claim on one currency funded by borrowing another, and the swap is the financing that settles the interest on both sides. The broker credits or debits your account at the daily rollover, usually 17:00 New York time, for every position still open (Dukascopy).
The credit or debit is the difference between the interest rate on the currency you bought and the rate on the currency you sold. Buy the higher-rate currency and the math pays you, sell it and the math charges you, which is the whole engine in one sentence.
The swap is not a broker fee dressed up as financing, despite how it sometimes feels, because the underlying mechanic is a genuine interest differential between two real central-bank rates. The broker's spread on the swap rate is where their margin sits, but the core number is the rate gap, not an invented charge (Dukascopy).
I treat the swap as the cost of carry, the price of holding the position through the night, and like any carry cost it only matters if you carry, which is the question that sorts day traders from everyone else.
The interest-rate differential
The engine of the swap is the interest-rate differential, and understanding it removes the mystery. Every currency has a central-bank interest rate, and a forex pair always has two, so holding a position means you receive the rate on the bought currency and pay the rate on the sold one.
If the bought currency's rate is higher than the sold one's, the differential is positive and the broker credits you each night. If it is lower, the differential is negative and the broker debits you, and the size of the credit or debit scales with the size of the gap.
The rates move when central banks move them, so a pair that paid a generous positive swap last year can pay nothing this year if the rate gap closed. The swap is not a fixed property of the pair, it is a moving function of two central-bank policies, and a holder who assumes last year's swap will persist is assuming today's rates will too.
I check the current swap rate on my broker's specification sheet before any multi-day trade, because the number that mattered at entry is the number today, and stale swap assumptions are a quiet way to misjudge a hold.
Positive versus negative swap
Whether you earn or pay depends on the direction you trade the pair, and the clean way to read it is which side carries the higher rate. A long position on a pair whose base currency has the higher rate typically earns positive swap, while the short on the same pair pays negative swap, and the mirror holds for the other direction.
The asymmetry is not always intuitive, because a pair can have you pay swap long and earn it short, or the reverse, depending entirely on which currency yields more. The rule is mechanical, you are paid for holding the high-yielder and charged for holding the low-yielder against it, regardless of whether you are long or short the pair overall.
The broker's specification sheet lists both the long and the short swap for every pair, often as a per-lot dollar amount, and reading that sheet is the one reliable way to know which side pays. Guessing from the direction alone gets it wrong often enough to matter over a long hold.
The detail the spec sheet often reveals is that, after the broker's own financing markup, both the long and the short swap on a pair can be negative at the same time. The interest-rate differential sets the raw number, but the broker adds a margin on each side, which can push both directions into debit territory on pairs whose rate gap is narrow, so there is no free side to hold on such a pair (broker spec sheet).
I never assume which side is positive, because the rate gaps shift and some pairs invert, and a thirty-second check of the spec sheet beats a month of wrong expectations about what a hold costs.
The Wednesday triple swap
The one swap detail that catches every new holder is the Wednesday triple charge, and it is worth stating plainly. Spot forex settles two business days forward, so a position held through Wednesday night covers the weekend, and the swap applied that night is tripled to account for Saturday and Sunday (Dukascopy).
The weekend days are not business days, so without the triple charge the financing for them would go unbilled, which is why the industry rolls them into Wednesday. The result is that holding over a Wednesday night costs or earns three times the usual daily amount, which can be a surprise on an account that expected the normal rate.
The practical effect is that a holder who is indifferent to a few nights of swap suddenly cares about Wednesday, because that one night carries half the week. A short-term trader who happens to hold over Wednesday pays a disproportionate slice of the week's carry on a single night.
I plan my holds around Wednesday when the swap is material, either closing before it to avoid the triple debit or holding through it deliberately when the trade's edge covers the cost, and treating it as a normal night is the rookie version of the error.
What swap costs, in numbers
The swap feels abstract until it is attached to a real lot and a real hold, and running the numbers once makes it concrete. The table below shows a representative negative-swap pair at a typical per-lot rate, held for different periods on a standard lot.
| Hold | Nights | Approx swap paid (1 std lot) | Read |
|---|---|---|---|
| 1 night | 1 | ~$7 | Trivial |
| 1 week (incl. Wed) | 5 | ~$49 | Noticeable |
| 1 month | ~22 | ~$154 | Material |
| 1 quarter | ~66 | ~$462 | Decisive |
The numbers are illustrative, because the exact rate varies by pair, broker, and the current rate gap, but the shape is universal. A swap that is trivial for one night becomes decisive over a quarter, and that compounding is what makes carry a real factor for any strategy that holds positions for weeks.
Flip the sign for a positive-swap pair and the same table becomes income rather than cost, which is the appeal that drives the carry trade in the section below. The magnitude is identical, only the direction of the money changes.
I size every multi-day hold with the swap totalled across the expected holding period, because a trade that looks profitable on the price move can be a loser once the accumulated carry is subtracted, and the table is the reminder that carry is never free.
Who pays swap and who does not
The cleanest split in forex is between traders who pay swap and traders who do not, and it maps almost exactly onto timeframe. A day trader who closes every position before the 17:00 rollover pays zero swap, because the position never exists overnight to be financed.
A swing trader who holds for days, a position trader who holds for weeks, and anyone running a multi-day strategy pays swap on every night the position stays open. The same trade, taken on the same setup, costs differently depending on how long it is held, which is why timeframe decides whether swap is a factor at all.
The implication for strategy choice is real, because a method that only works as a multi-day hold must clear its swap cost to be profitable, while an intraday method with the same gross edge pays no carry at all. The swap is one of the hidden reasons intraday trading has a structural cost advantage over holding, on top of the cost maths of the shortest timeframes.
I treat the rollover time as a hard line in my trading day, because crossing it deliberately is a decision to pay carry, and crossing it accidentally is a donation to the financing charge that the trade never budgeted for.
Swap and the hold strategy
Swap is the quiet enemy of any strategy that holds positions open for long periods, and naming the victims is useful. A grid strategy stuck in a trend holds its losing positions for days or weeks, and the swap bleeds the account on top of the floating loss, turning a recoverable drawdown into a structural one.
A swing trade that goes sideways, where price sits near the entry for a week, pays swap every night while producing no gain, so a trade that breaks even on price loses money on carry. A position trade that takes a month to reach target pays a month of swap, which must be subtracted from the profit to know the real result.
The honest consequence is that any hold strategy must generate enough directional edge to cover the accumulated carry, and a method that produces a thin gross edge can be a net loser once swap is counted. The grid trading page covers the strategy most exposed to this, because its stuck-in-a-trend positions sit open the longest.
I subtract the expected carry from every multi-day target before I judge the trade worthwhile, because a setup that pays ten pips over two weeks might pay nothing after the swap, and the gross-pip view hides that.
The carry trade, examined honestly
The carry trade is the strategy built around positive swap, and it deserves an honest look because its marketing is seductive. The method buys the high-yielding currency and sells the low-yielding one, collecting the positive swap every night as income, and on paper it looks like free money for waiting.
The catch is that the swap is a small daily credit while the currency risk is a large, discontinuous danger. A pair that pays a few dollars a night in swap can move hundreds of pips against the holder in a single risk-off session, and one such move erases months of collected carry in a day.
Carry trades collectively unwound in spectacular fashion during past risk shocks, because the same positions that earned steady swap were all closed at once when sentiment turned, and the rush for the exit turned a slow income into a fast loss. The strategy is correlation-sensitive, meaning many carry positions fail at the same moment, which is the opposite of diversification.
I do not call the carry trade wrong, because the positive-swap income is real and the method has paid for patient holders across calm years, but I treat the "free income" framing as the danger, because the income is small and daily while the risk is large and rare, and pricing only the income is how the rare risk ends an account.
Swap versus spread
Swap and spread are the two costs every forex trader pays, and confusing them leads to misjudging where the money goes. The spread is the per-trade cost paid on every position, open or closed intraday, while the swap is the per-night cost paid only on positions held past the rollover.
The spread is paid once at entry and once at exit, as the gap between bid and ask, and it is the dominant cost for high-frequency traders who take many trades and hold none. The swap is paid continuously, as a nightly financing charge, and it is the dominant cost for low-frequency holders who take few trades and hold them long.
The two costs favour opposite timeframes, which is the clean summary. The spread punishes the scalper and rewards the holder, while the swap punishes the holder and rewards the day trader, and the cheapest place to sit is the intraday middle where neither cost accumulates heavily.
I separate the two in my head because they demand opposite habits, minimizing spread means fewer trades at tighter pricing, while minimizing swap means closing before the rollover, and a strategy that ignores either pays it without realizing.
How to handle swap honestly
The honest approach to swap depends on your timeframe, and the rules are simple once the mechanic is clear. If you are an intraday trader, close your positions before the 17:00 rollover and swap ceases to exist for you, which is the cleanest possible solution.
If you are a holder, accept the swap as a real cost of the strategy and budget it into every trade's target, so a setup that cannot cover its accumulated carry does not get taken. Check the broker's spec sheet for the current long and short swap rates rather than assuming, because the rates move with central-bank policy.
If you run a hold strategy that gets stuck, like a grid in a trend, treat the accumulating swap as part of the decision to keep or close the position, because the carry turns a temporary drawdown into a permanent one the longer it sits. The sizing method that keeps any of this survivable is in the guide to volatility-based position sizing.
I handle swap by knowing my timeframe and acting accordingly, because the cost only bites holders, and a trader who knows whether they are one has already solved most of the problem.
Common mistakes with swap
The mistakes that swap causes are predictable, and naming them is most of the defence. The first is ignoring swap on a hold strategy, letting the nightly carry quietly erase a thin directional edge until a break-even trade becomes a loser.
The second is assuming the swap rate is fixed, trading on last year's positive-swap pair only to find the rate gap has closed and the credit is gone. The third is misjudging the side, guessing which direction earns and which pays instead of checking the spec sheet, and holding the wrong way for weeks.
The fourth is being surprised by Wednesday's triple charge, budgeting the daily rate and getting hit with three times it on the one night that covers the weekend. The fifth is the carry-trade error, collecting the daily swap income without pricing the rare currency move that can erase months of it in a session.
I keep the defence to one habit, checking the current swap on the spec sheet before any multi-day trade and budgeting it into the target, because that single action prevents most of the mistakes above and turns an invisible bleed into a cost I chose to pay.