Bid, Ask, and Spread: The Cost That Is Never Called a Fee

Ordering note. The v8 architecture listed this article as #4 and "Pips and lots" as #5. It is published as #5, since a spread cost cannot be expressed in money without pip value — see the pips and lots article, which supplies every figure used here; the swap was confirmed at QA and confirmed by Boss on 17 Aug 2026.

That difference between percentage-of-notional and percentage-of-margin is the whole subject of this article, and it is the figure the industry quotes least and a reader needs most. The previous article established pip size and pip value; this one turns them into money.

The two prices, and where the spread comes from

Every quote has two sides. The bid is the price at which your counterparty will buy the base currency from you; the ask (or offer) is the price at which it will sell to you. You always transact on the worse side for you — you buy at the ask and sell at the bid — and the gap between them is the spread. The mechanics are the same ones Pillar 6 established for any market; what differs in FX is scale and disclosure. The immediate consequence is that a position starts underwater. Open a position and close it instantly with no movement in the rate, and you lose the spread. There is no configuration in which this is not true, which makes the spread the one cost that is certain in advance while the outcome is not. Four things determine how wide it is. The pair. A heavily traded major carries a fraction of a pip; an exotic can carry tens of pips, and because pip size differs by quoting convention the money difference is larger than the pip counts suggest. The time. Spreads widen when liquidity thins — around session changeovers, in the hours when the relevant local market is closed, and at the week's open after the weekend gap. Events. Spreads widen sharply around scheduled data releases and central-bank announcements, and can widen dramatically in disorderly conditions. This matters more than it sounds: the moments a participant is most likely to want to act are the moments the cost of acting is highest. And the pricing model. Some platforms quote a variable spread that moves with market conditions; some quote a fixed spread, which is wider on average but predictable; and some quote a raw spread plus an explicit commission. None of these is free, and comparing them requires converting all of it into one number, which the broker article takes up.

Three costs beyond the spread, and the figure that matters

Slippage is the difference between the price you expected and the price you got. It arises because a quote is an indication until it is executed, and in fast conditions the executable price can be worse — sometimes materially. Slippage is not necessarily improper: a market that has moved has moved. But it is asymmetric in effect for a leveraged account, because it lands hardest in exactly the volatile conditions where positions are largest relative to margin. Financing, sometimes called swap or rollover, is charged or credited for holding a position past the daily cut-off, and reflects the interest-rate differential between the two currencies plus the platform's markup. Two consequences: a position can bleed money simply by remaining open, and a position can be credited for the same reason, which is what the carry trade attempts to exploit. Financing on exotic pairs with large rate differentials can be substantial in either direction. And conversion costs apply where the profit or loss arises in a currency other than the account's. Now the arithmetic the article exists for. A spread quoted in pips means nothing until it is converted into money, and money means little until it is measured against the capital at risk. Three steps: pips → money, using the pip value from the previous article; money → percentage of notional; then notional → margin, which is where leverage does its work. The general rule, worth memorising because it applies to every leveraged instrument and not just FX: leverage multiplies a percentage-of-notional cost by the leverage factor. A cost of 0.1% of notional is 3% of margin at 30:1, 5% at 50:1, and 20% at 200:1. The cost did not change; the base it is measured against shrank. That is why a spread described as "tight" can consume a large fraction of an account, and why the two figures — cost as a share of notional and cost as a share of margin — are not alternative presentations of the same fact but answers to different questions. The first describes the instrument. The second describes what happens to the reader.