What it is
Level 2 data is the resting order size sitting at each price level at a single moment. Rather than showing only the best bid and offer, it stacks quantity across a range of prices on both sides of the book, so you can see where resting interest is parked.
It is a photograph, not a recording. The moment it is taken, those orders can be repriced, pulled or filled. Level 2 tells you where resting interest sits right now; it does not tell you what that interest will do next. In futures, level 2 market data arrives as a depth-of-market feed, and how many levels you see and how often it updates depends on your feed and the contract.
- Top of book: best bid and offer only.
- Multi-level book: size at levels further out.
- Order-by-order data: every exchange-visible order from add to fill.
What it looks like in order-by-order data
Next to order-by-order data, level 2 looks like a summary. The same large number at a level is one event in the book and a whole sequence in the tape: add, cancel, add again, get filled.
That matters because identical resting size can mean different things. It may be genuine interest that is willing to trade, size that keeps getting hit and immediately refilled, or size that is pulled before price ever arrives. Only the order-by-order record separates those cases.
One naming trap: book imbalance is the buy-versus-sell imbalance of resting size at the ten-level depth. The per-price imbalance in a footprint is completed aggressive trading. Similar names, different things.
- Book imbalance: posted but not traded.
- Footprint imbalance: already traded, with direction.
Common misconceptions and limits
The first misconception is treating resting size as a commitment. Posted size is intent, not a promise; it can be pulled before price arrives or absorbed by aggressive flow. Icebergs are the extreme case: you infer one from a level being repeatedly refilled, and you can never confirm it.
The second is treating the book as static. Books flicker, and a snapshot can be stale before you finish reading it. A large resting order is not automatic protection either; it can be traded through, or pulled first.
The third is over-trusting book imbalance as a standalone signal. Whether you need level 2 at all depends on your approach: if your read requires knowing whether orders are actually willing to trade, whether liquidity is retreating, and whether size keeps refilling after fills, order-by-order data is the more direct input.
What the GC research says
The findings below come from TradeWhy research on COMEX gold futures (GC), data through 2026-08-31. They apply to GC only and should not be carried over to NQ or any other instrument.
Walls at key levels (W16, n=297, sample 2024-07 to 2026-08, ten-level book, held up on an independent period): divide wall size by the median ten-level size to get a relative size, then read the break probability within 30 minutes (%). Small walls 49, medium walls 48, large walls (1.3x and above) 31. In this GC sample, bigger resting walls showed a lower historical share of breaks within 30 minutes. That is a past-sample description, not a forecast. Its grade is single-period data, held up on an independent period: the result came mainly from one sample period and did not fall apart when re-tested on a separate one.
Ten-level book imbalance as a primary signal: this did not hold up in the GC data. It failed three independent re-checks and fits only a weak auxiliary role, worth roughly +8 percentage points. This is the imbalance of resting ten-level size, not the per-price aggressive imbalance in a footprint.
How to see it in TradeWhy
TradeWhy is an AI market-intent analysis and strategy validation tool built for CME Group futures day traders, starting with COMEX gold (GC) and CME Nasdaq (NQ). It puts real-time macro information, price response and L3 order-by-order behaviour side by side to explain what the dominant force is currently doing.
In that setup, level 2 plays the role of the photograph of currently resting size, while order-by-order behaviour answers the follow-up questions: whether orders are genuinely willing to trade, whether liquidity is retreating, whether size keeps refilling after fills, and whether heavy volume actually moves price. Reading the snapshot together with the process is more complete than watching the book alone.
Behaviour that repeats can be organised into strategies, which enter the strategy set only after historical and forward validation. TradeWhy explains market behaviour and helps users research and validate strategies; it does not give buy or sell recommendations or guarantee returns. The final trading decision is the user's.
FAQ
What is level 2 data in trading?
It is a snapshot of the resting bid and ask size stacked at each price level at one moment. Instead of only the best bid and offer, it shows quantity across a range of prices, often called the depth of the book. It describes where resting interest sits, not how much has traded.
Is there level 2 data for futures?
Yes. Futures contracts have depth-of-market feeds that show resting size across price levels. How many levels you can see and how often it updates depends on your feed and the contract, so the same instrument can look different on two setups.
Do I need level 2 data for futures trading?
That depends on your approach. A book snapshot shows only resting size and does not record the fill process. If your read needs to know whether orders are genuinely willing to trade, whether liquidity is retreating, or whether size keeps refilling after fills, order-by-order data is the more direct input. TradeWhy does not give buy or sell recommendations.
How is level 2 different from order-by-order data?
Level 2 is the picture of resting quantity at price levels; order-by-order data records each exchange-visible order from add and modify through cancel and fill. The same number at a level is one snapshot in level 2 and a sequence of events in the order-by-order record.
Do bigger resting walls mean price is less likely to break through?
In TradeWhy's GC sample, larger resting walls at key levels had a lower historical break rate within 30 minutes: small walls 49, medium walls 48, large walls at 1.3x and above 31, over 2024-07 to 2026-08. That is history, not a forecast, and it applies only to COMEX gold futures (GC), data through 2026-08-31.
Written by AI from TradeWhy's metric definitions, research findings and published product statements, then rule-checked. Research figures come from TradeWhy's study of COMEX gold futures (GC), data through 2026-08-31, and apply to GC only. Analysis only, not investment advice.