Understanding book imbalance

Use displayed bid and ask size as an observation to test.

The order book contains resting bids and asks. Those orders can trade, change or be cancelled, so displayed size is not a commitment to support a future price.

Check which part of the book is measured

The frozen Research book imbalance uses quantities at the best bid and best ask. It does not total every visible depth level. The reference owns the formula, valid-quote rules and a constructed example.

The terminal’s grouped DOM ladder exposes more depth and display choices. Do not substitute the visual weight of its full ladder for the top-of-book Research measurement.

Raw value, transformed rank and P-chip

Signed imbalance describes which side has more displayed top-of-book quantity. The separate imbalance rank feature transforms the magnitude using rolling statistics. A counted P-chip is another comparison against the symbol’s recorded values. These three numbers are not interchangeable.

Form a testable question

Choose a raw imbalance condition and scope in the Research Workbench. Inspect the eligible count, the baseline and both ordinary and dramatic matches. In replay, inspect whether the displayed size traded or disappeared. An imbalance alone does not identify spoofing, intent or future direction.