Skip to main content

The Order Book & Depth

Updated Jun 24, 2026 at 2:35pm

  • 1421320bdbde Level 1 vs Level 2 1 1,206
  • 1422a467cef2 Bid/Ask Stacking 1 1,211
  • 1420d263741d Spoofing & Iceberg Orders 1 1,229
Tree Key
Expandable — has sub-topics
475Local Id for node
a1b2c3d4Click to see full UUID
3Sub-topics
4Documents
5.0k wordsResearch depth
5Open node
Research Draft High 1,363 words

The order book is the live ledger of every resting (unexecuted) limit order for an instrument, organized into a bid side (buyers) and an ask side (sellers), with the quantity available at each price level. "Depth" is how much size sits behind the best prices — the stack of orders above the best offer and below the best bid that an aggressive order would have to consume to move price. This section covers what that book reveals, how traders read it, and — most importantly — how much of it lies. The core tension running through every topic here is that a resting limit order is a statement of intent, not a commitment: in modern electronic markets the large majority of displayed orders are cancelled before they ever trade, true size is routinely hidden, and a meaningful share of volume never appears on the lit book at all. So the order book is simultaneously the most granular window into supply and demand and one of the most adversarial, easily-misread data sources a trader can use.

What this section covers

The order book is the data structure underneath every electronic exchange. Orders are matched by a matching engine following price-time priority: better prices execute first, and among orders at the same price, the one entered earliest is filled first (it sits at the front of the queue). A limit order joins the book and adds depth; a market order (or a marketable limit order) crosses the spread and removes depth by hitting resting orders. The best bid and best offer define the inside market; the difference is the spread; everything stacked behind them is market depth or depth of market (DOM). This is verified standard microstructure (Wikipedia Central limit order book; SimTrade; Cube Exchange).

This section sits within Market Structure & Mechanics and is the microstructure-execution layer of the corpus. It is largely an intraday and execution domain — most of its content is consumed at sub-second to few-minute timescales by day traders, scalpers, market makers, and automated systems, not by position or swing traders. It defers the aggregate breadth/volume reading of supply-demand to other branches; here the unit of analysis is the individual resting order and the queue it sits in.

The core tension: displayed depth is not real depth

Three structural facts make the visible book an unreliable picture of true liquidity, and they organize the entire section:

1. Most displayed orders are cancelled, not filled. Hasbrouck & Saar (Low-Latency Trading) documented "fleeting orders" cancelled within milliseconds; later studies commonly cite eventual cancellation of roughly 95–96% of limit orders in U.S. equities (figure widely cited but study-dependent — see the child nodes). A displayed "wall" is statistically far more likely to vanish than to trade. 2. True size is deliberately hidden. Iceberg/reserve order types — offered natively by Nasdaq, NYSE, CME, and most venues — display only a small "tip" of a genuine large order, so a thin-looking level can conceal substantial real liquidity. 3. Much volume never reaches the lit book. U.S. equity trading is heavily fragmented and off-exchange: industry estimates put off-exchange (dark pools + wholesaler internalization) at roughly 40–50% of consolidated volume in recent years, with registered dark-pool ATSs a smaller slice (figures vary by source and definition; SEC officials have stated the bulk of retail marketable orders are internalized rather than routed to exchanges). The lit DOM therefore systematically understates the depth that actually exists.

The practical consequence — repeated across every child node — is that only executed volume (the time-and-sales tape) is hard evidence; displayed resting size is a noisy, partly adversarial signal that should confirm rather than lead.

Map of the sub-topics

  • Level 1 vs Level 2 (001-level-1-vs-level-2) — the data tiers. Level 1 is top-of-book (NBBO, best bid/ask + size, last trade) and is adequate for most investors and swing traders. Level 2 ("the ladder") adds multi-level depth and, on Nasdaq names, the market-participant ID behind each quote. Start here: it defines the raw inputs every other node reads, and clarifies that "Level 2" spans anything from a thin participant view to a full depth-of-book product (Nasdaq TotalView).
  • Bid/Ask Stacking (002-bid-ask-stacking) — the dynamic reading of depth: liquidity being added across price levels (stacking) versus withdrawn ahead of price (pulling), and the order-flow logic scalpers attach to each. Its key teaching is that the reliable signal is resting orders being consumed (absorption/exhaustion on the tape), not merely displayed.
  • Spoofing & Iceberg Orders (003-spoofing-and-iceberg-orders) — the two ways the book lies, on opposite sides of the law. Icebergs are legitimate concealment of real size; spoofing/layering is illegal placement of fake size with intent to cancel. This node carries the enforcement record (Coscia, Sarao, JPMorgan's $920.2M CFTC settlement) and the honest caveat that real-time spoof detection by a screen-watching trader is unproven — "big order appeared then vanished" is consistent with spoofing but also with ordinary cancellation or iceberg lag.

Read together, the three move from what you can see (L1/L2) to how it behaves (stacking) to why it deceives (spoofing/icebergs).

When it matters vs. when it doesn't

Order-book depth matters most for execution quality on thin or fast-moving names, for very-short-horizon order-flow trading in deep centralized limit-order-book markets (liquid futures like ES/NQ/CL, some crypto), and for routing and working large orders. Academic microstructure work confirms a genuine edge in order-book imbalance — but at seconds-to-tens-of-seconds horizons captured by automated systems, not by a human eyeballing a ladder.

It matters far less, and is often misleading, for swing and position trading: at multi-day horizons the intraday book is invisible and irrelevant, displayed walls flicker faster than any discretionary decision, and the off-exchange/hidden share means the lit picture is incomplete. For these traders, Level 1 fields (spread, displayed size at the inside, volume) are a sufficient liquidity/slippage sanity check; raw depth reads add noise, not edge.

Strengths & limitations

The order book's strength is unmatched granularity — for execution, queue position, and short-term liquidity assessment, nothing else shows the actual supply-demand stack. Its limitations are structural and permanent: it is partial (hidden/iceberg/dark liquidity invisible), ephemeral (orders pull in milliseconds), fragmented (one feed rarely shows the whole market), and adversarial (spoofing and layering exploit the very belief that walls are firm). The single most common misuse across the whole section is treating the displayed book as the complete book and reacting to large resting orders as firm commitments — the orders most likely to move price are often the ones you cannot see, and the visible walls are the cheapest to fake.

Sources