Why the advertised spread is a best case, not your cost
An advertised spread is a marketing figure drawn from a venue's own best conditions: the most liquid instrument, the calmest hours, the top account tier. It is usually the minimum or a flattering average, not the spread you get at the moment you click. Your real cost is whatever gap exists between bid and ask on your instrument, at your size, in the second your order is filled.
The spread on a pricing page is the narrowest number a venue can honestly print: the best instrument, the calmest hours, the top account tier, often the lowest reading rather than the usual one. The spread on your fill is whatever the gap between bid and ask happened to be in the second your order reached the book, at your size, on your instrument.
Both numbers are real. They just describe different events. Understanding why they drift apart tells you roughly how far apart they will be the next time you look.
The spread exists because someone is taking the other side of your trade
Every quote you see is two prices: the highest price anyone currently offers to buy at, and the lowest price anyone currently offers to sell at. The gap between them is the spread. You buy at the higher one and sell at the lower one, which means that at the instant you enter a position, you are already behind by the width of that gap.
The gap is not arbitrary. The participants posting those prices are agreeing to trade against whoever shows up, without knowing whether the next arrival knows something they do not. If the price is about to move, they get filled at a price that is immediately stale. The spread is what they charge for carrying that risk, plus whatever competition between quoters compresses it down to.
That single fact explains most of what spreads do. When the next price is easy to predict, the risk of being picked off is low, competition dominates, and the spread tightens. When the next price is hard to predict, the risk rises and the quote widens. Volatility, thin hours, session changeovers and scheduled announcements all push the same direction.
Minimum, average and typical are three different claims
A pricing page usually gives you one of three things, and they are not interchangeable.
A minimum is the best reading observed, and it is what the word “from” is doing when it appears in front of a number. It is the lower edge of the distribution. It is accurate and it is not a forecast of your fill.
An average is the mean across some window. This sounds fairer and often is not, because the mean is weighted by time, not by trading. A market is quoted around the clock and traded unevenly. The hours when spreads are narrowest tend to be the hours when the fewest orders arrive. An average over all hours therefore pulls toward conditions almost nobody traded in.
A typical or median figure is the middle of the distribution: half of readings above, half below. It is the most useful of the three and the rarest to see published, because it is never the flattering number.
| What is quoted | What it describes | How it relates to your fill |
|---|---|---|
| Minimum (“from X”) | The narrowest reading observed | A floor you will rarely touch |
| Average | Mean across a time window | Dragged down by quiet hours |
| Typical / median | The middle of the distribution | Closest to a fair expectation |
| Your own fills | The spread you actually paid | The only sample that is about you |
There is a second layer underneath all three: which instrument, which account tier, and which size. An advertised spread is normally quoted on the single most heavily traded instrument the venue offers, at the top tier, for an order small enough to fill entirely against the best price on the book. Change any of those and the number changes with it.
A worked example of the gap
All inputs here are invented for the arithmetic. They are not what anything charges.
Suppose a venue advertises a spread of 0.4 units, quoted as a minimum. Say you trade a less active instrument, where the usual gap is 1.2 units. Say the moment you trade is a session changeover, and the gap is 2.0 units then.
Imagine one unit of spread costs you $1 per contract, and say you trade 5 contracts per round trip.
- At the advertised minimum: 0.4 x $1 x 5 = $2.
- At the usual gap on your instrument: 1.2 x $1 x 5 = $6.
- At the changeover moment: 2.0 x $1 x 5 = $10.
Now say you place 20 round trips a month. On the advertised figure that is $40. On the realistic figure for your instrument and your timing, somewhere between $120 and $200. The advertised number was off by a factor of three to five, and nothing dishonest happened anywhere in the chain. Every assumption behind the quoted number was simply different from yours.
Then add size. Suppose the best price on the book is good for 2 contracts and you send 5. Two fill at the top, and the remaining three fill at worse prices further down. Say that costs an extra 0.5 units averaged across the order: another 0.5 x $1 x 5 = $2.50 per round trip, which is more than the entire advertised spread. That extra portion is slippage, and it is invisible on a pricing page because a pricing page quotes a price, not a quantity. You can carry this kind of arithmetic through with your own figures using the crypto fee calculator.
The mistake is treating one number as a comparison
The common error is not believing the advertised spread. It is lining up advertised spreads from several venues side by side and treating the smallest one as the cheapest venue.
That comparison fails because the numbers are not measured the same way. One venue publishes a minimum, another an average, a third an average on a different instrument or a different tier. Comparing them is comparing differently defined quantities that happen to share a unit. The venue with the smallest printed number may have the widest typical spread of the group.
Three habits fix most of this.
- Find out which statistic you are reading. If the page does not say whether it is a minimum, an average or a median, treat it as a minimum, because that is what an unlabelled figure almost always is.
- Check which instrument and which tier the figure refers to, and whether either describes you. A headline spread on the most traded instrument says little about the one you actually hold.
- Compare total cost per round trip, not spread alone. A narrow spread with a commission attached can cost more than a wider spread with none, and the only way to know is to add them at your own size. The comparison pages and the glossary set out how the components fit together.
The second mistake is subtler: assuming the spread is a fixed property of a venue at all. It is a property of the market in that moment, which the venue passes on and sometimes marks up. On a quiet afternoon in a heavily traded instrument you may get close to the advertised figure. In the first minute after a surprise, you will not, at any venue, and no amount of shopping around changes that. Maker and taker fees and fee tiers behave more predictably, because they are published schedules rather than live market conditions.
So the practical reading of an advertised spread is narrow: it tells you the best case this venue is willing to claim, under conditions it chose and may not have disclosed. It is a floor, not an estimate. Your own fill records, mid price at send versus price at execution, are the only measurement of the spread that is actually about your trading.
Education only, not financial advice. Trading carries a risk of loss.
Key facts
- The spread is the gap between the best available buy price and the best available sell price, so you pay it the instant you enter, before the position moves at all.
- An advertised figure is almost always a minimum ('from X') or an average, and the two describe very different experiences.
- A simple average can be dominated by the quiet hours, when spreads are narrowest and almost nobody is trading.
- Advertised spreads are typically quoted on the single most liquid instrument and the best account tier, neither of which may apply to you.
- Spreads widen with volatility, at session open and close, around scheduled events, and for orders larger than the quoted size at the top of the book.
- Your own fill statements are the only sample of the spread that is actually about you.
Frequently asked
Is a wider fill than the advertised spread a sign of something dishonest?
Usually not. A quoted minimum is a real number that was real under the conditions it was measured in, and quoting a minimum is standard practice. The gap between it and your fill comes from conditions: your instrument, your size, your account tier, and the moment you traded. It becomes a problem only when a venue presents a minimum as if it were typical, with nothing on the page to say otherwise.
Why do spreads widen at exactly the moments I want to trade?
Because the people quoting prices are taking on risk, and that risk is highest when the next price is hardest to predict. During a news release or a session changeover, a market maker who quotes a narrow spread can be filled at a price that is stale a second later. Widening the quote is how that risk gets priced. The moments that look most attractive to trade are the moments the spread costs most.
Should I prefer a zero-spread or raw-spread account?
Those labels move the cost rather than removing it. A raw or zero-spread account typically passes through a narrower market spread and charges a separate commission per lot or per side, so the total is spread plus commission. Compare total cost per round trip at the size you actually trade, not one component of it.
How do I estimate my own typical spread without special tools?
Record the mid price you saw when you sent the order and the price you were actually filled at, for a few dozen trades at the times you normally trade. The distribution of those differences is your spread, including slippage. The median of that set is more useful to you than any figure on a pricing page.
Risk warning: trading is high risk and most retail accounts lose money. Education only, not financial advice.