Spread, Depth and Slippage: What a Market Order Costs

Three numbers behind one fill

Every market order pays two costs that are easy to conflate and should not be.

• The spread is the distance between the best bid and the best ask. It is a property of the top of the book, and it does not depend on your size.
• The depth is how much size is resting at each price away from the top. It is a property of the whole book, and it decides how your size behaves.
• The slippage is your order’s own result: the difference between the best available price and the average price you actually got, once your size has walked through however many levels it needed.

The first is a constant you can read before you trade. The second is the reason the third is not.

What we measured, and when

To keep this from being a rule of thumb, the numbers below come from one snapshot of Binance’s public order-book endpoint, taken at 00:17 UTC on 2026-10-15, reading 1,000 price levels on each side for three USDT pairs. The endpoint is public and requires no key, so the measurement is repeatable - the values, obviously, are not.

At the top of the book:

The one-cent spread is identical on all three, and that is not a coincidence: all three pairs carry a tickSize of 0.01, which is the smallest price increment the venue accepts for them. One tick is therefore the tightest spread any of these markets can show.

What differs is what a cent is worth. The same tick is roughly 0.0012 basis points of an 81,774 price and roughly 0.96 basis points of a 104 price - a factor of about 800. So the headline “the spread is one cent” tells you nothing usable until you divide it by the price.

Walking the book

A market buy does not get one price. It consumes the best ask, then the next, and so on until it is filled. Simulating that walk against the snapshot gives this:

Slippage here is measured against the best ask at the moment of the snapshot, in basis points, so it is the additional cost on top of the spread you already pay by crossing it.

Four things in that table are worth reading carefully.

Small orders paid nothing. A 1,000 or 10,000 USDT market buy filled entirely at the best ask on all three pairs. The best level held more size than the order needed, so there was no second level to walk to. If your orders are this size, depth is not what is costing you money - the spread is.

The first measurable slippage appeared at very different sizes. On BTCUSDT, 100,000 USDT still filled at the best ask. On ETHUSDT it needed a second level and cost 0.0058 basis points. On SOLUSDT it also needed a second level and cost 0.2597 - about 45 times more for the same notional.

The ranking held as size grew, but the ratio did not. At 1,000,000 USDT the cost was 1.60, 2.58 and 8.18 basis points: SOLUSDT cost about five times what BTCUSDT did. At 5,000,000 the spread widened to 4.99, 9.13 and 156.00 - SOLUSDT now cost about 31 times BTCUSDT. Slippage does not scale linearly, and the difference between two markets does not stay proportional either.

The books ran out. A 20,000,000 USDT market buy could not be filled within the 1,000 levels the endpoint returned on any of the three pairs. That is a limit of the measurement, not a claim that no such order could execute - the visible book is a view, and liquidity arrives as the order is worked.

Why the same order size behaves differently on two pairs

Two different mechanisms were on display, and separating them is the useful part.

The spread difference is arithmetic. Same tick, different price. Nothing about the SOL market is worse because its spread is 0.96 basis points; the tick is simply a larger fraction of a smaller number.

The slippage difference is structural. It comes from how much size rests at each price level, and how far apart those levels are in relative terms. A step of one cent is 0.0000122% of an 81,774 price and 0.00957% of a 104 price - so moving even one level away from the top costs far more, proportionally, on the cheaper asset. The deeper reason a 5,000,000 USDT order on SOLUSDT consumed 454 levels while the same notional on BTCUSDT consumed 453 levels and still cost 31 times less is that each of those levels was a much larger relative step.

That is the practical warning. A pair can look deep - SOLUSDT had about 9.67 million USDT of asks within the 1,000 levels, comparable to BTCUSDT’s 8.88 million - and still cost a large order far more, because depth measured in dollars does not tell you the shape of the ladder.

What this means for the order you place

Three habits follow from the numbers rather than from theory.

Measure before sizing, not after. The slippage table above takes seconds to reproduce for a pair you are about to trade, using the venue’s public depth endpoint. The number that matters is your order size against that pair’s actual ladder - not against a different pair, and not against a figure from an article.

Treat the spread and the slippage as separate line items. If your orders are small relative to the top level, your cost is the spread and it is predictable. If they are not, the spread is the smaller half of the bill and it is the less important half.

Re-measure when the market changes. These are single-instant readings. Order books thin out in fast markets and thicken in quiet ones, and the snapshot that told you 100,000 USDT was free on BTCUSDT was a statement about one moment at 00:17 UTC.

The limit order is the answer, and it has its own cost

The obvious response is to use a limit order, and it does remove both costs in the table: a limit order cannot fill worse than its price. But it replaces them with a different set of costs, and they are real.

• It may not fill. Resting at or behind the best price means waiting for someone to cross the spread and trade with you.
• The market may leave. While the order rests, the price may move away, and the fill you were waiting for may arrive only in the direction you did not want.
• It may be filled in a way you did not plan. Partial fills leave you with a position that is neither what you intended nor nothing at all, which matters if the order size came from a risk calculation.
• Your rate changes with your role. Adding liquidity and removing it are charged differently, and the trading fee calculator shows what that difference is worth on a given size - though on a tight pair like BTCUSDT, the 1.60 basis points of slippage at 1,000,000 USDT is of the same order as the fee difference between the two roles.

Which cost matters more is not a general question. It depends on whether the fill or the price is the thing you cannot do without, and no measurement in this article answers that for you.

For why a limit order might be refused before it ever rests, see why an exchange rejects your order. For the maker and taker distinction behind the fee difference, see maker and taker orders. The position size calculator turns a risk budget into a quantity, which is the input these measurements need. Where the two venues differ in structure is covered in the Binance profile and the Binance vs OKX comparison.

Risk note. The figures in this article are one snapshot of Binance’s public order-book endpoint taken at 00:17 UTC on 2026-10-15, and order books change continuously. Reproduce the measurement for the pair and size you intend to trade rather than relying on these values. Basis points of slippage are a transaction cost, not the main risk in a trade; nothing here is investment advice, and a fill at a good price can still be a losing position.