Book Pressure

Quick Reference

Property	Value
Dimension	signal
Category	order_flow
Version	v0.9.0 (Beta)
Output Column	book_pressure

Book pressure: sum(level_weight * level_imbalance for level in orderbook_levels) - weighted sum across levels

Formula

CDM Inputs

Column	CDM Table	Description
cdm_lob_snapshot.bids	cdm_*	CDM source table
cdm_lob_snapshot.asks	cdm_*	CDM source table

Parameters

Parameter	Type	Default	Description
max_levels	integer [1, 50]	10	Maximum number of levels to consider
weight_decay	float [0.0, 1.0]	0.9	Decay factor for level weights (higher levels get lower weights)

Output

Column: book_pressure

Weighted sum of level imbalances across order book

Market Intuition & Trading Rationale

Book pressure computes a depth-weighted imbalance across order book levels: Σ(level_weight × level_imbalance). Unlike simple depth_imbalance (which weights all levels equally), book pressure applies exponentially decaying weights — the touch (level 1) gets the highest weight, level 10 gets the lowest. This captures an important microstructure reality: imbalance near the mid-price is more predictive of short-term price moves than imbalance deep in the book.

The multi-step computation extracts bid/ask sizes at each level, computes per-level imbalances, generates a decay weight vector, and produces the weighted sum. This is an array-heavy feature — in DolphinDB Community Edition, it requires the consolidated LOB handler pattern (parsing JSON once, computing all LOB features in a single row loop).

Positive book pressure means more bid liquidity than ask liquidity, weighted toward the touch — buyers are positioning aggressively near the mid. Negative means sellers are positioned. The magnitude indicates conviction: strong pressure with tight clustering near the touch is higher quality than weak pressure spread across many levels.

Usage Cases

• Pre-trade direction signal: Positive book_pressure → bullish (more bids near the touch). Trade in the direction of pressure. This is a leading signal — liquidity providers position before aggressive traders arrive.
• Spoofing resistance: Weighted pressure is harder to manipulate than simple depth imbalance. A spoofer placing a large order at level 10 gets minimal weight; genuine positioning at the touch gets maximum weight. The decay weights act as an anti-spoofing filter.
• Execution entry timing: Execute buys when book_pressure is positive — there's ample bid liquidity to absorb your sell. Execute sells when negative. Your passive orders are more likely to fill when resting liquidity is deep on the opposite side.

YAML Definition

name: book_pressure
description: 'Book pressure: sum(level_weight * level_imbalance for level in orderbook_levels)
    - weighted sum across levels'
category: order_flow
version: v0.9.0 (Beta)
dimension: signal
status: Pre-release
required_inputs:
  - cdm_lob_snapshot.bids
  - cdm_lob_snapshot.asks
output_column: book_pressure
output_description: Weighted sum of level imbalances across order book
tags:
  - book_pressure
  - orderbook
  - weighted
  - imbalance
  - microstructure
  - complex
parameters:
  max_levels:
    type: integer
    description: Maximum number of levels to consider
    required: false
    default: 10
    constraints:
      min: 1
      max: 50
  weight_decay:
    type: float
    description: Decay factor for level weights (higher levels get lower weights)
    required: false
    default: 0.9
    constraints:
      min: 0.0
      max: 1.0
steps:
  - primitive: TRANSFORM
  op: array_extract
  params:
    array: cdm_lob_snapshot.bids
    field: size
    output_col: bid_sizes
  - primitive: TRANSFORM
  op: array_extract
  params:
    array: cdm_lob_snapshot.bids
    field: price
    output_col: bid_prices
  - primitive: TRANSFORM
  op: array_extract
  params:
    array: cdm_lob_snapshot.asks
    field: size
    output_col: ask_sizes
  - primitive: TRANSFORM
  op: array_extract
  params:
    array: cdm_lob_snapshot.asks
    field: price
    output_col: ask_prices
  - primitive: TRANSFORM
  op: array_subtract
  params:
    output_col: level_imbalances
    a: cdm_lob_snapshot.bids
    b: cdm_lob_snapshot.asks
  - primitive: TRANSFORM
  op: generate_weights
  params:
    n_levels: $max_levels
    decay: $weight_decay
    output_col: level_weights
  - primitive: TRANSFORM
  op: weighted_sum
  params:
    values: level_imbalances
    weights: level_weights
    output_col: book_pressure