What a forex spread really costs you
Spreads widen in a clean order of liquidity. EUR/USD trades tightest at roughly 0.1 to 1.2 pips, the other dollar majors between 0.5 and 2.5, the crosses 1 to 5, and exotics such as USD/TRY can exceed 20 pips, so the pair you choose sets your cost before your strategy does.
I treat the spread as a market-set commission rather than a broker fee, because that is exactly what it is. It is the gap between the bid and the ask, set by liquidity, and you pay it twice on every trade.
On a standard lot each pip on EUR/USD is worth about $10, so a one-pip spread is $10 round-turn before any profit (ForexEAShop). That is why the currency pair you choose often matters more than your strategy: the most volatile pair is not the cheapest to trade if it charges you twenty pips to enter and exit.
Spread comparison by currency pair: the typical ranges
The table below sets out those typical ranges in full. Read each band as a floor rather than an average, because live spreads only touch the low end during peak liquidity.
I draw the figures from live aggregator data (Myfxbook) and broker-published typical spreads, with raw ECN pricing at the low end and standard accounts at the high end.
| Pair | Type | Typical spread | Round-turn / lot |
|---|---|---|---|
| EUR/USD | Major | 0.1 - 1.2 pips | ~$2 - $12 |
| USD/JPY | Major | 0.2 - 1.5 pips | ~$2 - $15 |
| GBP/USD | Major | 0.5 - 2.0 pips | ~$5 - $20 |
| AUD/USD, NZD/USD | Major | 0.7 - 2.5 pips | ~$7 - $25 |
| EUR/GBP, EUR/CHF | Cross | 1.0 - 3.0 pips | ~$10 - $30 |
| EUR/JPY, AUD/JPY | Cross | 1.5 - 4.0 pips | ~$15 - $40 |
| GBP/JPY | Cross | 2.0 - 5.0 pips | ~$20 - $50 |
| USD/MXN | Exotic | 15 - 50 pips | ~$150 - $500 |
| USD/TRY, USD/ZAR | Exotic | 20 - 80+ pips | ~$200 - $800+ |
Read the low end as raw ECN pricing during peak liquidity. IC Markets publishes an average EUR/USD spread of 0.1 pips on its raw account, which is why it keeps topping low-spread broker tests (IC Markets, 2026).
Read the high end as a standard account during normal hours. BestBrokers' 2026 broker testing averaged 0.86 pips on EUR/USD across the field, a useful midpoint for what a typical retail account actually pays (BestBrokers, 2026).
The dollar majors, pair by pair
The seven dollar majors are where spreads live or die for most retail traders, so they deserve a pair-by-pair read. Each one earns its spread band from a specific liquidity profile, and I trade each at the session where its book is deepest.
EUR/USD is the benchmark, typically 0.1 to 1.2 pips, and it carries roughly a fifth of daily global forex volume (Audacity Capital). It is the pair I default to whenever execution cost matters more than the headline move.
USD/JPY sits a hair behind at 0.2 to 1 pip, with the second-deepest order book in forex (TradingBrokers). Its spread stays tight through both the Asian and New York sessions, which is rare.
GBP/USD, known as cable, trades 0.5 to 2 pips, a touch wider because sterling's book is thinner than the euro's. I find it cheapest during the London morning when UK banks are adding liquidity.
USD/CHF runs a similar 0.5 to 2.5 pips and often mirrors EUR/USD in reverse. Its spread tightens whenever euro liquidity is strong, since the two books share so many participants.
The commodity dollars sit together at roughly 0.7 to 2.5 pips. AUD/USD is cheapest in the Asian session when Sydney and Tokyo are both open, while USD/CAD trades tightest at the London to New York overlap.
NZD/USD is the widest of the majors at about 1 to 3 pips, because the kiwi book is the thinnest of the seven. I treat it as a majors-priced pair that occasionally trades like a cross during quiet hours.
These seven are the G10 majors, and they share one trait that matters here: deep enough order books that the spread rarely wrecks a reasonable stop. The full 28-pair majors list extends the same logic across the most liquid combinations.
Where minors and crosses get expensive
Crosses like EUR/GBP and EUR/JPY trade real volume, but they lose the dollar anchor that tightens the majors. Without two deep dollar sides, the book thins and the spread widens.
EUR/GBP typically runs 1 to 3 pips, while EUR/JPY sits closer to 1.5 to 4. The yen crosses carry more volatility, so their spreads reflect that risk premium as well as the thinner book.
GBP/JPY is the expensive one, normally 2 to 5 pips, because the yen side narrows the pool (TradingBrokers). When a UK or Japanese release surprises, I have seen GBP/JPY spreads push out by another 5 to 8 pips before price settles.
Spread-to-range ratio matters more than the raw number here. A 3-pip spread on a pair that averages 150 pips a day is cheap, while the same 3 pips on a 30-pip daily mover is brutal.
Exotics are the spread trap
Exotic pairs such as USD/TRY and USD/ZAR look exciting on a chart, and that is exactly how they trap traders. Their spreads routinely run 20 pips or more, which is a 20-pip loss the moment you click buy (ForexEAShop).
On a standard lot, a 20-pip spread is about $200 gone at entry, compared with about $5 for a 0.5-pip EUR/USD spread (ForexEAShop). I avoid them for intraday trading and only use them, rarely, for long swings where the spread is a small fraction of the expected move.
The order books are thin and the currencies carry high political and economic risk, so spreads widen fast and stay wide (MondFx). That structural thinness is the real reason, not any single news event.
Fixed and variable spreads price different risks
Brokers quote either fixed or variable spreads, and comparing them directly is a category error. A fixed spread holds steady through normal conditions while a variable spread floats with the market.
Fixed spreads run wider because the broker carries the risk of holding the price during volatility. EUR/USD fixed spreads sit around 1.3 to 5 pips against 0.1 to 3 pips variable (Audacity Capital).
I prefer variable spreads almost always, which is why they dominate 2026 retail pricing. The trade-off is that they spike exactly when you want to trade, around news, and a 0.1-pip quote can become 5 pips in seconds.
The full cost stack: spread, commission and swap
The spread is only the first of three costs on a forex trade. Commission and overnight swap complete the bill, and a tight spread can hide a worse deal overall.
A raw ECN account might quote 0.1 pips but charge around $3.50 per side in commission, about $7 round-turn per lot. Against a 0.1-pip spread worth roughly $1, the commission is the real cost, not the spread (IC Markets pricing).
Swap, the overnight financing charge, bites positions held past the New York close. I check the swap rate before holding any trade more than a session, because a few nights of negative swap can exceed the spread you saved.
Session timing changes the spread you actually pay
The same pair trades at very different spreads across the 24-hour cycle, because liquidity follows the world's financial centres. The London to New York overlap, roughly 13:00 to 17:00 GMT, is the tightest window of the day.
The Tokyo session thins the book, so yen crosses widen even though the market is open. I plan entries around the overlap whenever the setup allows, because that is where spreads sit closest to their published floor.
Friday close and Sunday open are the worst liquidity windows. Spreads gap wide as the book empties, and holding through the weekend adds swap risk on top.
What widens a spread beyond the pair itself
The pair sets the floor, but conditions set the actual number you pay. The same EUR/USD that trades at 0.1 pips at the London to New York overlap can gap to 3 pips at the Tokyo open.
Three forces widen spreads, and I rank them by how often they catch traders. Low-liquidity sessions come first, then broker model, then high-impact news, with news the one that actually blows up accounts because the spread spikes as stops get hunted.
Broker model is the factor most traders miss. A raw ECN account shows a tiny spread because the commission is billed separately, often around $7 round-turn per lot, while a standard account bakes the cost into a wider spread.
The all-in cost is what determines your stop placement, not the headline spread.
In 2026, competitive pressure and zero-commission marketing have compressed headline majors spreads further, which is why average tested EUR/USD spreads now sit near 0.86 pips (BestBrokers). The catch is that the saving has migrated into commission and overnight swap, so compare round-turn, not the banner number.
What a spread does to a real trade
Run the spread over a hundred trades and the pair choice becomes the strategy. A scalper taking 100 trades on EUR/USD at 0.8 pips pays about 80 pips in spread over the run, roughly $800 on a single standard lot (ForexEAShop).
The same 100 trades on an exotic at 20 pips pays 2,000 pips, about $20,000, before a single winning trade is booked. I have watched new traders blow accounts on exotics without realising the spread was doing the damage, not the market.
This is why the cheapest-looking pair is usually the right one for high-frequency strategies. The edge in scalping is thin enough that a single extra pip of spread can flip a profitable system into a loser over a month of trading.
How to compare spreads without getting fooled
Compare round-turn cost in dollars, never the raw pip number. A 0.1-pip raw spread plus $7 commission is a worse deal than a 0.8-pip standard spread with no commission below a certain trade size.
I use a live aggregator like Myfxbook to read actual spreads at the session I trade, because broker-published typical figures are best-case, peak-liquidity numbers. A gold, XAU/USD, spread of 0.3 points in testing can be 3 points the moment CPI prints.
Standardised broker testing closes the gap between marketing and reality. CompareForexBrokers' 2026 methodology converts every spread to a standard USD cost across six majors, which is how IC Markets and Pepperstone keep trading places at the top (CompareForexBrokers, 2026).
The discipline that protects you is simple. Pick the pair for the setup, verify the live spread at the moment you trade, then size the cost against your stop.
Do that and the spread stops being a tax and starts being a variable you manage.