Order Book Imbalance Edge

The quant playbook for trading order book imbalance. Iceberg detection, queue position, and the micro-edge that prop firms pay six figures to learn.

Order book fundamentals

• Limit order book mechanics: maker/taker fees, time priority, pro-rata vs. FIFO matching
• What 'imbalance' actually means: bid-ask depth ratios, top-of-book vs. cumulative depth, and the right look-back window
• Queue position: how to estimate it without seeing it, and why it dominates HFT PnL

Imbalance signals that work

• L1 imbalance: 5-second persistence as a directional predictor — base rate by instrument and time of day
• L5 vs. L10 depth: when extending the book improves signal and when it adds noise
• Trade-flow imbalance: the difference between resting and aggressive volume and how to combine them

Iceberg and hidden liquidity

• Detecting icebergs in real time: 4 statistical tells using only public market data
• How to size your trade around an iceberg: stepping ahead vs. fading the level
• Dark-pool prints and how to read TRF data without paying for direct feeds

Implementation

• Python notebook with three signal implementations (depth imbalance, trade-flow imbalance, queue-position estimator) using public NASDAQ TotalView samples
• Realistic backtesting pitfalls: queue priority assumptions, fill rate haircuts, latency-arbitrage exclusions
• How to ship a 'manual quant' strategy: discretionary execution informed by quant signals, not a full HFT stack