Order Book Depth: Why Your Fill Is Worse Than the Price

The price on your screen is the best available price for a small trade, not the price your trade will actually get. Order book depth is the measure of how much size sits behind that headline number, and it decides whether your order fills near it or somewhere much worse. A great many retail losses that feel like bad luck are really depth problems: the quote was genuine, but it was genuine for a tenth of the size you sent.

Understanding order book depth needs one idea and a little arithmetic. The quoted price is the top of the book. Your fill is the average of every level your order eats through on the way down. When the book is deep, those two numbers sit close together. When it is thin, they can be far apart — and the gap widens faster than almost anyone expects as order size grows.

• The quoted price is the top of the book; your average fill is what you actually paid, and they are rarely the same number.
• Walking the book level by level gives you the real average, and slippage grows non-linearly as size increases against fixed depth.
• Iceberg orders and hidden liquidity mean the depth you can see is not always the depth that exists.
• A limit order does not remove the problem — it converts slippage into fill risk, which is a different cost.

What order book depth actually measures

Depth measures the total quantity of resting orders at each price level away from the best bid and ask. It is not a single number you look up; it is a shape. A book can be deep at the touch and hollow two levels out, which is the configuration that hurts most retail orders.

The top of the book is not your price

The best ask is simply the cheapest unit currently offered. It says nothing about what the next thousand units cost. Screens that show only one price — the last trade, or the mid — hide this completely, which is why the first market order on a new pair so often surprises people.

Think of it as a queue at a shop with one item left at the advertised price. The sign is honest. It just was not written for you.

Why the average fill is the only number that matters

Your realised cost is a weighted average across every level you consumed. If you buy through three price levels, you did not pay the first one — you paid a blend, and the blend is what your position is measured against. Profit, loss and break-even all get calculated from the average, never from the headline.

This is also why two traders buying “the same price” on the same day can hold positions at noticeably different costs. One sent a small order into the touch. The other sent a large one through it.

Walking the book: a worked example

Walking the book means summing the cost of every level your order touches, in order, until the size is filled. Here is a sell-side ladder for a thin pair, with a market buy of 500 units.

1. Level one: 100 units offered at 10.00. Cost so far: 100 × 10.00 = 1,000.00.
2. Level two: 150 units offered at 10.05. Cost: 150 × 10.05 = 1,507.50. Running total: 2,507.50.
3. Level three: 250 units offered at 10.12. Cost: 250 × 10.12 = 2,530.00. Running total: 5,037.50.
4. Fill complete at 500 units. Average price = 5,037.50 ÷ 500 = 10.075.
5. The screen said 10.00. The fill was 10.075, about 0.75% worse — before any fee.
6. Now double the order to 1,000 units. The next level offers 500 units at 10.25: cost 5,125.00. Total 10,162.50, average 10.1625 — about 1.63% worse.

Double the size, more than double the damage. That is the non-linearity: the first few levels are cheap, the ones behind them are not, and every extra unit pushes you further into prices nobody advertised. On a genuinely thin pair the same 1,000-unit order can fill several percent away from the quote, which on a leveraged position is the difference between a plan and a liquidation.

Depth is a promise made by strangers, and every promise in the book can be withdrawn the moment it would cost the maker money to keep it.

How order book depth fails when it matters most

Depth fails precisely when volatility rises, because the orders that made the book look safe are the first to be pulled. Market makers widen or vanish; the ladder that looked solid at lunch is three levels of thin quotes during a news candle.

Iceberg and hidden orders

An iceberg order displays a small tip while the rest of the size refills behind it at the same price. A book showing 200 units at a level may in fact hold 5,000. This cuts both ways: hidden size absorbs your order with less slippage than the screen implied, or it disappears the instant it is filled, and you learn nothing either way. Large venues also allow non-displayed liquidity that never appears in the visible ladder at all.

Depth that is displayed but not real

Some displayed size exists to influence what you think, not to trade. Orders are placed, cancelled and replaced faster than a human can read. If a wall of bids evaporates the moment price approaches it, the depth was decoration. You cannot reliably distinguish this from ordinary market-making in real time, which is the honest reason to treat visible depth as a hint rather than a guarantee.

Reading order book depth before you size a trade

Reading depth is a pre-trade check, not a forecast. You are estimating cost, not direction.

Limit orders convert slippage into fill risk

A limit order caps your worst price, which solves slippage and creates a new problem: you may not trade at all. That is the honest trade-off, and it is a trade, not a free upgrade.

If you place a limit at 10.00 in the ladder above and price runs to 10.12, you keep your 0.75% and hold nothing. In a market that then moves without you, the opportunity cost exceeds the slippage you avoided. Both outcomes are real costs; the difference is that one is visible on your statement and the other is not.

A practical middle path is to size against the depth you can see. If your order is a small fraction of the visible size within a tight band of the touch, slippage is likely to be modest. If it is larger than the first few levels combined, you are not really trading at the quoted price at all — you are trading at whatever the book happens to contain that minute, and you should decide that deliberately rather than discover it in the fill notification.