Helen is Marketing Team Lead at Motion Trade with 4+ years in Web3 and crypto marketing. Before joining Motion Trade, she built and led the marketing function at CLS Global and managed social media campaigns for a portfolio of crypto clients at Ninja Promo. She specializes in turning complex trading products into clear stories.
Order book depth is the amount of money resting in the book within a defined distance of the mid price. The most commonly published version is ±2% depth, which sums the value of buy orders between the mid price and two percent below it, and sell orders between the mid and two percent above. It answers a question that volume cannot: how much can be traded right now without moving the price much. Spread describes the cost of a small trade, depth describes the cost of a large one, and volume describes what already happened rather than what is currently possible. Exchanges monitor all three, and depth is the one a project can improve deliberately.
Depth is the least understood of the three numbers on a token's market page, partly because it is the only one that is not self-explanatory from its name and partly because different data providers measure it at different distances from the mid price and rarely say so prominently.
This is what the number is, how it is computed, why the measurement band matters more than the figure, and how to work out what your own token needs.
What an order book contains
An order book is the list of unfilled limit orders at a venue. Buy orders, called bids, sit below the current price. Sell orders, called asks, sit above it. The highest bid and the lowest ask are the top of the book, the distance between them is the spread, and the midpoint between them is the mid price, which is the reference every depth calculation uses.
Nothing requires any particular price level to be occupied. The book is only populated where somebody has chosen to rest an order, which is why a market with few participants has visible gaps and a market with many does not. The structure is covered in more detail in our explainer on order books and spread.
How ±2% depth is calculated
The calculation is arithmetic rather than modelling. Take the mid price, define a band two percent above and below it, and add up the notional value of every resting order inside that band on each side separately.
Suppose a token's best bid is 0.0100 and its best ask is 0.0102, giving a mid price of 0.0101. The upper band runs to 0.010302 and the lower band down to 0.009898. Every sell order priced at or below 0.010302 counts toward the +2% depth, and every buy order priced at or above 0.009898 counts toward the -2% depth. Orders further out contribute nothing to the figure, however large they are.
| Metric | What it measures | What it misses |
|---|---|---|
| Spread | The cost of trading a small size immediately, as the gap between best bid and best ask | Says nothing about size; a tight spread can sit on top of an almost empty book |
| ±2% depth | Money resting within two percent of the mid price, on each side separately | Ignores everything outside the band, and treats a wall at the edge the same as depth spread evenly |
| 0.1% and 1% depth | Liquidity very close to the price, which is what execution actually consumes first | Harder to find published for smaller tokens, since fewer providers show it |
| 24h volume | What was traded over the past day | Is historical, can be inflated, and says nothing about current available size |
| Slippage on a fixed order size | The realised cost of a specific trade, which is the number a buyer actually cares about | Depends on the size chosen, so it is only comparable when the size is held constant |
Why the measurement band matters more than the number
Two percent is a convention rather than a law, and it became the default largely because CoinGecko and CoinMarketCap publish it. CoinGecko's trust score for a market pair combines the order book spread and ±2% depth with volume and traffic measures. CoinMarketCap takes a different approach with a liquidity score derived from the slippage a trader would incur across a range of order sizes, which is closer to what a buyer experiences.
Professional data providers frequently measure closer in. Kaiko's research notes that its most common depth measures are 0.1% and 1%, and that it moved away from 2% depth specifically because that level is more frequently gamed, precisely because it is the level reported on public sites. That is worth sitting with. A book can be arranged to look strong at the published band while being hollow inside it, and the arrangement is visible to anyone who measures at a tighter band.
The practical conclusion is that depth is only meaningful alongside the band it was measured at, and that a project assessing its own liquidity should look at both a tight band and a wide one. If the 2% figure is healthy and the 0.1% figure is negligible, the book is decorative.
What depth looks like on large assets
Scale is useful for calibration. CoinGecko's liquidity research across eight major exchanges found a median cumulative depth of roughly twenty to twenty-five million dollars on each side within one hundred dollars of the bitcoin price, with Binance accounting for around a third of it. Ether showed roughly fifteen to sixteen million dollars of depth within a tenth of a percent of its price.
Those figures are not a target for a growth-stage token. They are a reminder that depth on a mature asset is measured in tens of millions within a very narrow band, while depth on a newly listed token is frequently measured in tens of thousands within a much wider one. The gap explains why the same size order behaves completely differently in the two markets.
Depth, spread and volume are not substitutes
A tight spread with no depth behind it is the most common misleading configuration. The top of the book shows a narrow gap, the market page looks healthy, and then a trade of any real size falls through several empty levels and executes far from the last print. This is the mechanism behind most complaints of manipulation on small tokens, and it is described in what happens to a token without a market maker.
Volume is the weakest of the three as an indicator of market quality, because it is historical and because it is the metric that manufactured activity targets. Depth and slippage are harder to fake convincingly, since they require capital to be actually at risk in the book rather than merely reported. That asymmetry is why exchanges and data providers moved toward liquidity-based metrics in the first place, and it is a large part of why fake volume now ends listings rather than sustaining them.
What exchanges measure, and what they do about it
Listing venues monitor these numbers continuously, and at least one publishes the thresholds. MEXC's ST Warning Rules name an average daily buy-sell spread above two percent for fifteen consecutive days as grounds for a warning tag, alongside a fifteen-day average price deviation above fifteen percent against other centralised exchanges, fewer than one hundred holders with more than five dollars of the token, and total holder balances averaging below fifty thousand USDT daily for thirty consecutive days. Where the risk is assessed as severe, delisting can follow three days after the tag.
Most venues do not publish equivalent numbers, but all of them monitor the same properties, and the character of the process is the same everywhere: measured daily, averaged over consecutive-day windows, and therefore not repairable retroactively. The broader picture is in why tokens get delisted.
How much depth does a token need?
There is no universal figure, and any desk that quotes one without asking about your float is guessing. The inputs that determine it are the day-one or current circulating float, the valuation, the number of venues the liquidity has to cover, and the largest realistic sell event in the period rather than the average day.
The useful way to set a target is to work backwards from an acceptable execution outcome. Decide the largest order a normal participant might reasonably place, decide how much price impact you are willing for that order to cause, and size the depth so that the impact stays inside it. That converts an abstract number into a testable commitment, which is exactly what belongs in a market making contract rather than a promise about volume. The full sizing method is in how much liquidity a token needs at TGE, and the contractual form it should take is in what belongs in a market making agreement.
Reading your own book in ten minutes
Open the token's market page on a data provider and look at the spread and ±2% depth for each listed pair. Note where the depth is concentrated: if one venue carries almost all of it, the other listings are contributing distribution rather than liquidity, and the price on those venues will drift.
Then open the exchange's own order book and look inside the band. Count the levels between the mid price and one percent out, and see whether they are populated evenly or whether the depth sits as a single block near the edge of the measurement band. Even distribution absorbs trades smoothly; a block at the edge produces a jump as soon as anything eats through the near levels.
Finally, check the price of the same pair across every venue where it trades. A persistent difference means arbitrage is not operating, usually because the spread makes it unprofitable or the size is too small to be worth the capital, and cross-venue divergence is one of the conditions monitoring systems watch explicitly.
FAQ
What is order book depth?
It is the total value of resting limit orders within a defined distance of the mid price, calculated separately for the buy and sell sides. It measures how much can be traded immediately without moving the price beyond that distance, which is a different question from how much has been traded historically.
What does ±2% depth mean?
It is the sum of the value of buy orders priced between the mid price and two percent below it, and separately the sum of sell orders priced between the mid and two percent above it. Orders outside that band are excluded from the figure regardless of size.
What is a good ±2% depth for a token?
There is no absolute answer, because the meaningful figure scales with float and valuation. The practical approach is to define the largest order a normal participant might place, decide the maximum acceptable price impact for that order, and size depth so the impact stays within it.
What is the difference between spread and depth?
Spread is the distance between the best bid and the best ask, so it describes the cost of trading a small amount immediately. Depth is the money available within a band around the mid price, so it describes how much can be traded before the price moves. A tight spread on a shallow book is common and misleading.
Is depth a better measure than trading volume?
For assessing whether a market can absorb a trade right now, yes. Volume is historical and is the metric that manufactured activity most often targets, while depth requires capital to be genuinely resting in the book. This is why data providers and exchanges increasingly weight liquidity metrics over reported volume.
Why do providers measure depth at different percentages?
Because different participants care about different distances. Short-horizon traders care about liquidity within a fraction of a percent, longer-horizon holders care about a wider band. Kaiko notes that it favours 0.1% and 1% and moved away from 2% because that level is more often gamed, given that it is the level published on the major public sites.
Can order book depth be faked?
It can be arranged to flatter a specific published band, typically by concentrating orders near the edge of the two percent band while leaving the levels closest to the price empty. It is harder to fake than volume, because the capital has to actually rest in the book, and the arrangement becomes obvious as soon as depth is measured at a tighter band.
How does depth work on a decentralised exchange?
An automated pool has no order book, so depth is a function of the pool's reserves and its pricing curve. There is a quote for every size at some level of impact, which removes the empty-book problem, but the pool does not adjust its quotes as the market moves and cannot vary depth by conditions. The comparison is set out in CEX versus DEX market making.
Do exchanges have minimum depth requirements to stay listed?
Most do not publish numeric minimums, but all monitor liquidity conditions continuously and act on sustained deterioration. MEXC is the clearest example of published criteria, naming a two percent average daily spread over fifteen consecutive days among its warning triggers. Elsewhere the criteria exist without being disclosed, and the review process is the first visible sign of them.
