Crypto order book explained: bids, asks, depth and walls
Crypto order book explained: bids and asks, spread, market depth, slippage, buy and sell walls, spoofing. Worked example from the Binance BTC book, Nov 4, 2026.

Contents
- What an order book is made of: bids, asks and the spread
- The spread: the price of immediacy
- Market depth: how many BTC it takes to move the price
- How to read the order book in five steps
- Buy walls and sell walls: support or bait
- Spoofing: a wall placed in order to be pulled
- Order book, trade tape and volume together
- Common mistakes
- FAQ
A crypto order book is the live list of every limit order waiting to be filled on an exchange: bids on one side (how much buyers will pay, and for how many coins), asks on the other (how much sellers want). The gap between the best bid and the best ask is the spread, and the volume resting at each price level is market depth. Reading the order book means answering three questions: what an immediate trade costs, how much size the market can absorb before the price moves, and whether a large resting order is real or bait. Below is how the book works on Binance and Coinbase, what the BTC/USDT book looked like on November 4, 2026, and why a sell wall is not the same thing as resistance.
What an order book is made of: bids, asks and the spread
Every centralized exchange matches trades through a book of resting orders. Limit orders that do not fill immediately queue up by price, then by time: Coinbase Exchange documentation states that orders are filled based on price-time priority. The best price goes first, and at the same price the earlier order wins. A market order, per the same page, executes immediately against available liquidity and always pays taker fees. How the different order types fit into this queue is covered in our guide to crypto order types.
Binance serves its book through the /api/v3/depth endpoint. The documentation says the response holds a bids array and an asks array, each entry a price and quantity pair, plus a lastUpdateId. The default is 100 levels and the maximum is 5,000, and a 5,000-level request costs 250 units of rate limit against 5 for the first hundred. One thing the book never contains is trades. It shows resting intentions only; executed trades are a separate feed.
| Term | What it is | What to look at |
|---|---|---|
| Bid | Resting buy order | Best bid is the most anyone will pay right now |
| Ask | Resting sell order | Best ask is the least anyone will sell for |
| Spread | Best ask minus best bid | The cost of an immediate round trip |
| Depth | Total size resting within a band around the price | How much the market absorbs before moving |
| Wall | One level with an unusually large size | Whether it holds or vanishes as price approaches |
The spread: the price of immediacy
The spread is the first cost you pay for speed. A market buy lifts the ask, a market sell hits the bid, and the difference goes to the market before any fee. On liquid pairs it is tiny. In a snapshot of the Binance BTC/USDT book we pulled on November 4, 2026 at 3:07 AM ET through that endpoint, the best bid sat at $78,407.55 and the best ask at $78,407.56: a $0.01 spread, one tick. Coinbase BTC-USD showed $78,260.06 against $78,260.07 in the same minute, also one cent. Bybit spot BTC/USDT showed $78,394.2 against $78,394.3.
The gap between venues matters more than the spread here: Binance and Coinbase were almost $150 apart at the same moment. That is not free arbitrage but different quote currencies (USDT versus USD) and different pools of buyers. How that gap turns into an indicator of US demand is explained in our article on the Coinbase premium.
Market depth: how many BTC it takes to move the price
A tight spread does not mean a deep market. Depth is the volume resting in a band around the price, which is to say the market's ability to absorb a large order without moving: the greater the depth, the smaller the impact of a market order on the price. Kaiko measures liquidity exactly that way, as the volume available within 1% of the best price, and writes that the higher an order's size, the higher its price impact and slippage, because the order consumes one level after another until it is filled.
Here is what the Binance BTC/USDT book showed at 3:07 AM ET on November 4, 2026 (5,000 levels per side, our calculation from the depth endpoint):
| Band around mid price | Bids, BTC | Bids, $ | Asks, BTC | Asks, $ |
|---|---|---|---|---|
| ±0.25% | 111.4 | 8.7M | 198.4 | 15.6M |
| ±0.5% | 133.6 | 10.5M | 234.5 | 18.4M |
Five thousand bid levels reached down to $77,232.92 (1.5% below the price), while five thousand ask levels reached only $79,132.95 (0.93% above): sell orders were packed more densely, and the 5,000-row cap did not even cover one percent on the upside. That is practical lesson number one: a single request is not enough to measure "2% depth"; you need the full book or a data provider such as Kaiko.
Slippage from the same snapshot: a market buy of 1 BTC would have filled at an average of $78,412.54, 0.006% worse than the best ask; 10 BTC at 0.021% worse; 50 BTC at 0.04%; 100 BTC at 0.062%, roughly $49 per coin. A market sell of 50 BTC would have cost 0.027% off the best bid. For bitcoin that is noise; for a thin altcoin the same size can cost whole percentage points. Kaiko's piece cites August 5, 2024, when BTC-USD slippage on major US exchanges tripled within hours and one exchange's BTC-EUR pair saw slippage above 5%.
How to read the order book in five steps
- Find the middle. Best bid, best ask, spread between them. A spread wider than normal for the pair means either a thin market or a sharp move that just happened.
- Check the size on the first rows. On November 4 the best bid held 7.87 BTC and the best ask only 0.19 BTC. Buying size at the top price was impossible; the order would walk up the levels.
- Compare both sides over the same band. Asks within ±0.5% were almost double the bids (234.5 BTC versus 133.6 BTC). That is called an imbalance, yet BTC was up 1.7% over 24 hours per our dashboard at 3:00 AM ET. Imbalance describes the stock of orders, not the flow of trades.
- Spot the walls. On the sell side: $78,800 (82.67 BTC) and $78,500 (54.66 BTC). On the buy side: $77,600 (8.06 BTC) and $77,500 (4.72 BTC). All four are round or "psychological" numbers where humans and algorithms like to park orders.
- Test the wall against the tape. A wall that absorbs market orders on contact is real. A wall that disappears when the price gets within a few dollars was bait.
Buy walls and sell walls: support or bait
Cointelegraph's guide defines a buy wall as a massive buy order, or multiple buy orders, around a particular price level, and a sell wall as a significant accumulation of sell orders at a given price. On a depth chart they look like vertical steps, and traders read them as support and resistance. The same article warns that orders can be pulled or introduced at any time and that large holders can use their capital to create or remove walls as a way to manipulate the market.
Hence the rule: a wall is informative only through its behavior. The $78,800 order on Binance on November 4 was almost ten times larger than any bid in the same band. If the price reached it and the trade tape showed a seller actually handing out dozens of BTC, that is resistance. If it vanished $20 before contact, it was a decoy. One snapshot cannot tell the two apart.
Spoofing: a wall placed in order to be pulled
Placing orders you do not intend to fill has a legal name. In the United States, Commodity Exchange Act section 4c(a)(5)(C), as amended by Dodd-Frank, prohibits conduct that "is, is of the character of, or is commonly known to the trade as, spoofing (bidding or offering with the intent to cancel the bid or offer before execution)". The CFTC's 2013 interpretive guidance lists four examples: overloading the quotation system, delaying another person's execution, "submitting or cancelling multiple bids or offers to create an appearance of false market depth", and creating artificial price movements. It adds that even a single instance can be a violation and that intent beyond recklessness is required.
The first case under that provision was settled on July 22, 2013: Panther Energy Trading and its principal paid $2.8 million over an algorithm that placed a small sell order it wanted filled, then several large buy orders at rising prices to suggest demand, and cancelled them as soon as the small order executed. In crypto the same mechanic looks like a wall that appears and disappears every time the price approaches. Regulation of centralized crypto exchanges differs by country, so the safe assumption is that spoofing in the book is possible, and that a wall should be judged by its behavior rather than its size.
Order book, trade tape and volume together
The book shows intentions; the tape shows facts. A resting order can be cancelled, a printed trade cannot. Experienced readers watch three things at once: the book (what is resting), the tape (what has filled) and daily volume (how much has traded in total). Binance streams a partial book of 5, 10 or 20 levels every second or every 100 ms, plus a separate diff stream for maintaining a local copy of the full book.
A useful pairing is volume against depth. Kaiko proposes a volume-to-depth ratio: when traded volume far exceeds resting liquidity, the market is fragile. Per our numbers at 3:00 AM ET on November 4, 2026, BTC traded 14,483 BTC, about $1.12B, over 24 hours, with the Fear and Greed Index at 34 ("fear"). Set that against $18.4M of asks within 0.5%: daily volume exceeded visible depth by a factor of dozens. That is normal in crypto, and it is why large orders get sliced and why sharp moves cascade into futures liquidations.
Liquidity is also uneven across the clock. In its research on the ETF effect on spot markets, Kaiko notes that the weekend share of BTC volume fell to 16% in 2024 from a 28% peak in 2019, and that the US market close on weekdays accounts for over 6.6% of all volume on benchmark exchanges. The same book on a Sunday night is thinner than on a Tuesday afternoon in New York.
Common mistakes
- Reading a wall as a level. A wall is an order, not a trade. Until the tape shows fills, it is a hypothesis.
- Comparing depth across exchanges from one request. Binance returns up to 5,000 levels, Bybit returned 200 in our call, Coinbase 500. A "±2%" band covers a different slice of each book.
- Assuming imbalance predicts direction. On November 4 asks were double the bids and the price still rose over the day. Imbalance measures stock, not flow.
- Watching spot and ignoring derivatives. For BTC, turnover and positioning are concentrated in futures: per our data at 3:00 AM ET on November 4, open interest stood at $7.67B with funding at 0.01% per 8 hours. What that means is covered in our guide to open interest.
- Sending a market order into a thin pair. If an altcoin's book is thinner than your order, you create the move you then have to live with. A limit order or slicing solves it.
FAQ
What is an order book in simple terms? A list of every buy and sell order waiting to be filled, with a price and size for each. The exchange matches them by best price, then by time of arrival.
How is the order book different from a depth chart? The book is a table of levels; the depth chart is the same levels added up cumulatively, bids curving to the left, asks to the right, with walls showing as steps.
Can you predict price from the order book? No. The book shows resting orders that can be cancelled. It helps you estimate the cost of a trade and slippage risk, not direction.
Where can I see the bitcoin order book for free? In any exchange's interface or through public APIs: Binance returns up to 5,000 levels without an API key. Live prices and market indicators are on our bitcoin page.
This is not investment advice.



Comments