Early access is open. Free sample datasets are live now. Download a sample

Learn

Crypto Order Book Depth and Slippage

7 min read · Updated 19 November 2026 · By the marketdata.co data team

The price on the screen is only good for the first few coins. Anything bigger eats into the book, and the fill you get depends on how much size sits near the mid. This page shows how to read depth data and turn it into a slippage estimate before you send the order.

Details

What depth bands tell you

A full L2 order book lists every price level and its size. That is a lot of data, and most of the time you only need one answer: how much can I trade within X% of the mid price? Depth bands answer it directly. Each row gives the cumulative size and notional resting on one side of the book, out to a given distance from mid.

Here is one real snapshot from our free BTC perpetual sample, taken at 00:00:01 UTC on 1 November 2026:

distance_from_mid_pct side cumulative_size cumulative_notional_usd
-5 bid 10138.132 771074735.2
-2 bid 6015.362 461632878.88
-1 bid 2817.967 217317556.75
-0.2 bid 522.381 40450511.3
0.2 ask 589.675 45760508.91
1 ask 2682.932 209014668.23
2 ask 5818.703 455408530.67
5 ask 7090.01 556870511.04

Source: Binance. All rows have symbol = BTC-USDT-PERP and timestamp = 2026-11-01T00:00:01Z; the -4, -3, 3 and 4 bands are omitted here for space.

Reading it

Within 0.2% of mid there was about $45.8 million on the ask side. A $10 million market buy fits well inside that band, so its worst fill would be less than 0.2% above mid, and its average fill much closer. An order of $150 million would not fit within 0.2% and would reach into the 1% band, where $209 million sits.

Look at the shape as well as the size. At 5% the bid side held about $771 million against $557 million on the ask. Out at the edges, buyers had more size waiting than sellers. Near the mid it was the reverse: the ask had slightly more inside 0.2%. Imbalances near the mid change minute to minute; the far bands move more slowly and say more about where large resting interest sits.

Code

Estimate slippage in Python

The function below interpolates linearly between bands. That assumes size is spread evenly inside each band, which is rough but good enough to size orders.

"""Example: estimate worst-case and average slippage for a market order."""
import numpy as np
import pandas as pd

book = pd.read_csv("btc-perp-depth-bands.csv", parse_dates=["timestamp"])
snap = book[book["timestamp"] == book["timestamp"].min()]

def slippage(snap, side, notional_usd):
    s = snap[snap["side"] == side].copy()
    s["dist"] = s["distance_from_mid_pct"].abs()
    s = s.sort_values("dist")
    dist = np.concatenate([[0], s["dist"].values])
    cum = np.concatenate([[0], s["cumulative_notional_usd"].values])
    if notional_usd > cum[-1]:
        return None  # order is larger than the visible book
    worst = np.interp(notional_usd, cum, dist)
    return worst, worst / 2  # average is about half the worst under even spread

for size in [1e6, 1e7, 1e8]:
    print(f"${size:,.0f} buy:", slippage(snap, "ask", size))

For a $10 million buy on this snapshot, the worst level comes out at about 0.044% from mid, and the average fill about half that. At $100 million it is roughly 0.47%. The jump is the point: slippage grows much faster than order size once you leave the tightest band.

Watch out

Common mistakes

Using one snapshot as the rule. Depth changes all day and thins out around news and on weekends. Run the estimate across many snapshots and size orders to the thin end, not the average.

Trusting displayed depth fully. Some resting orders are cancelled the moment price approaches them. Treat far bands as an upper bound on what will really be there.

Ignoring fees. Taker fees are paid on top of slippage. On a small order the fee can be larger than the price impact.

Comparing coin sizes across assets. 500 BTC and 500 ETH are very different amounts. Compare depth in cumulative_notional_usd when you look across pairs.

Details

Get the data

Free BTC perpetual depth sample. 300 rows of cumulative depth bands from 0.2% to 5% each side of mid, with size in coins and notional in USD.

FAQ

Frequently asked questions

What is order book depth in crypto?

It is the total size of resting buy and sell orders at each price, or within a distance of the mid price. More depth near the mid means a market order moves the price less.

How do I calculate slippage from order book data?

Walk up the book from the best price, filling your order against each level until it is complete. The last price you reach is your worst fill, and the size-weighted average of all levels is your average fill. Depth bands give you a quick approximation of the same result.

Why does slippage grow faster than order size?

Size near the mid is limited, so larger orders reach levels further away. Each extra dollar fills at a worse price than the last.

Is perpetual depth the same as spot depth?

No. Perps and spot trade in separate books, often with different depth. Check depth on the exact market you plan to trade.

Keep reading