交易的真实成本：点差、隔夜利息、佣金——完整账单

Every trade pays a bill, and the bill arrives whether the trade wins or loses. Most of this industry would rather talk about anything else — which is exactly why this sheet itemizes every way a CFD trade costs money, then computes the full bill of one ordinary trade before it is placed. A broker that itemizes its bill is making a checkable claim. That is the standard to hold anyone to, including us.

The four line items

Strip away the account tiers and the pricing pages, and a CFD trade can cost money in exactly four ways. Every fee any broker charges on a trade is one of these wearing different clothes:

• The ▲spread — the gap between the buying and the selling price. Paid once, on entry, by every trade without exception; it is why a new position opens slightly negative.
• The ▲commission — a fixed fee per trade, charged on some instruments (share CFDs, typically) and some account types, usually per side: once to open, once to close.
• The ▲swap — overnight financing, debited (occasionally credited) for every night a position stays open. Zero for trades closed the same day; the dominant line for trades held weeks.
• The ▲conversion fee — charged when a trade's profit or loss is settled in a currency other than your account's, on its way into your balance.

One distinction keeps beginners' accounting honest: costs are not losses. A loss is the market disagreeing with your position — uncertain, sometimes large, sometimes absent. Costs are certain, capped, and payable in both outcomes. You cannot avoid them; you can only know them, and knowing them before the trade is the entire skill this sheet teaches.

Each line lives somewhere findable. The spread is on the quote panel right now — the distance between the two prices. Commission rates sit on the account-type page, per instrument. Swap rates are in a published schedule, updated as interest rates move, and visible per instrument inside the platform. Conversion fees are in the same schedule, as a percentage. Four numbers, four known addresses, zero excuses for meeting any of them for the first time on a statement.

The worked bill: three nights, one lot, fully costed

Take the survey's standard trade and run the whole bill. A USD account buys 1 standard lot of EUR/USD at 1.0850 — $10 per pip — with a 0.8-pip spread, a $3.50-per-side commission on this account type, and an illustrative published swap of −$5.40 per night for long EUR/USD. The position is held three nights and closed. Every figure here is illustrative — the live ones belong to the published schedules — but the method transfers to any pair, any size, any broker: that is what it is for.

Now set the bill against both possible worlds, because it sits in both. The trade closes 50 pips up: gross +$500, net +$468.80. The trade closes 50 pips down: gross −$500, net −$531.20. Same bill, both directions — trimming the win and deepening the loss by the same $31.20. And read it as a hurdle: at $10 per pip, the position must move roughly 3 pips in your favor just to pay for itself. Every trade you ever place starts that far behind, which is worth knowing before you place thirty of them in a week.

Two scheduling details complete the picture. Swap is applied at the daily rollover — 5pm New York — so a position held across that moment pays a night even if it was opened an hour before. And one night each week, usually Wednesday, is charged at triple rate to account for weekend settlement. Neither detail is a trick; both are in the published schedule, and the deep-dive sheet on overnight financing derives exactly where the numbers come from.

The fourth line stayed at zero only because the example used a USD account on a USD-quoted pair. Change one assumption and it wakes up: the same trade settled into a EUR account passes its $500 result through a currency conversion, and a 0.5% conversion fee turns that into about $2.50 — charged on the loss as well as the profit, because the conversion happens either way. Small on one trade; a permanent percentage skimmed off everything, across hundreds. Traders who deal mostly in pairs quoted in their account currency never meet this line. Traders who don't should know its rate by heart.

Why “zero commission” never means “zero cost”

Commission is the only line item brokers can remove and still run a business, because the spread can carry the bill instead. That is the entire anatomy of the “zero commission” offer: not a discount, a relocation. An account with raw spreads might quote EUR/USD at 0.2 pips plus $7 commission — total $9 on our worked lot. A commission-free account might quote the same pair at 1.2 pips — total $12, no commission line anywhere, and a slightly larger bill folded invisibly into the entry price.

Neither structure is dishonest, and which one is cheaper for you depends on trade size and frequency. Run both shapes across a month to see it: thirty 1-lot trades cost $270 on the raw-spread account ($9 each) and $360 on the commission-free one ($12 each) — the active trader saves $90 by paying the visible fee. Two trades a month and the difference is $6, in the other structure's favor once smaller typical sizes are figured in. The structure should be chosen with your own arithmetic, not the pricing page's favorite adjective.

What deserves suspicion is the framing, not the structure: a page that says “zero commission” is telling you about one of four line items and hoping you hear “free.” The test never changes — compute the all-in cost of one specific trade under each structure, with published numbers, and compare totals. Marketing compares adjectives; statements compare dollars.

Cost per style: scalps pay spread, swing trades pay swap

Which line dominates your bill is decided by how you trade, not by the broker. A scalper taking ten 1-lot trades a day pays the 0.8-pip spread ten times: $80 per day, before any trade has won or lost, against targets of perhaps 8 pips each. For that style, the spread is not a detail — it is the largest opponent, met on every single attempt, and a half-pip difference in average spread changes the arithmetic of the whole approach.

A swing trader holding one position for 20 nights pays the spread once — $8 — and the swap twenty times: $108 at our illustrative rate, plus the triple Wednesdays. The spread that decides the scalper's month barely registers; the financing line the scalper never sees becomes the whole bill. A day trader who is flat by each rollover sits between the two: spread and commission, never swap.

Scale matters as much as style. The scalper's $80 a day compounds to roughly $1,600 over a twenty-day trading month — on a $10,000 account, 16% of the capital paid in spread alone, before the strategy has been right or wrong even once. That arithmetic is not an argument against frequent trading; it is the entry fee frequent trading must clear, and most people quoting their win rate have never computed it. The practical consequence: read your own statement, not the pricing page. The statement shows which lines your actual behavior pays, in dollars, over a real month — and most traders are surprised by what dominates.

The habit: cost it before you place it

Everything above compresses into a habit that takes under two minutes per trade and pays for itself indefinitely:

1. Read the live spread off the quote panel and multiply by your pip value — entry cost.
2. Add the commission for your account type, both sides.
3. Estimate your holding period; if it crosses any rollover, multiply the published swap rate by the nights, counting the triple night.
4. Add the conversion percentage if your profit will settle in another currency.
5. Set the total against your target: if the bill is a meaningful fraction of what the trade aims to make, the trade is thinner than it looked.

A broker that publishes every input to that checklist is letting you audit it — which is the only version of “low cost” worth believing. A broker that makes any input hard to find has answered a different question, and answered it clearly. The bill always exists either way. The only choice you get is whether you read it before the trade or after.