---
title: Shallow versus Deep: The Anomaly in the Limit Order Book
url: https://www.ml-quant.com/papers/ssrn/7675440/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-10-30
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: SSRN 7675440
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7675440
featured: 2026-10-30
citations: unknown
topic: Trading, Microstructure & Execution
---


# Shallow versus Deep: The Anomaly in the Limit Order Book

Shallow-book imbalance shocks predict higher future volatility and positive returns, while deep-book shocks predict lower future volatility and negative returns, a pattern robust across Chinese stock constituents since 2014.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7675440
- Identifier: SSRN 7675440
- Released: 2026-10-27
- First featured: Quant Letter No. 137 (2026-10-30): https://www.ml-quant.com/issues/2026-10-30/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: Trading, Microstructure & Execution
- Authors: Lin Fu, Wuyi Ye, Keli Wang

## Related

- [Painting the market: generative diffusion models for financial limit order book simulation and forecasting](https://www.ml-quant.com/papers/arxiv/2509.05107/): A new method for simulating financial market data transforms limit order book data into an image format and uses diffusion models to predict future states, showing top performance on LOB-Bench.
- [Bimodal dynamics of the artificial limit order book stock exchange with autonomous traders](https://www.ml-quant.com/papers/arxiv/2508.17837/): The paper uncovers the inherent bistability and complex dynamics of an artificial stock market exchange, which emerge from micro-level trading rules.
- [Detecting Multilevel Manipulation from Limit Order Book via Cascaded Contrastive Representation Learning](https://www.ml-quant.com/papers/arxiv/2508.17086/): The study suggests a learning framework to enhance the detection of trade-based manipulation in financial markets, with Transformer-based architectures proving most successful.
- [Reinforcement learning for trade execution with market and limit orders](https://www.ml-quant.com/papers/arxiv/2507.06345/): The paper presents a reinforcement learning framework for optimal trade execution, using multivariate logistic-normal distributions, which outperforms traditional benchmark strategies.
- [ClusterLOB: enhancing trading strategies by clustering orders in limit order books](https://www.ml-quant.com/papers/arxiv/2504.20349/): ClusterLOB is a new method for grouping individual market events in financial markets, aiding in understanding market structure and participant behavior, and developing effective trading signals.
- [Learning the Spoofability of Limit Order Books With Interpretable Probabilistic Neural Networks](https://www.ml-quant.com/papers/arxiv/2504.15908/): The article discusses a new real-time detection model for spotting potential market manipulation in cryptocurrency exchanges, with 31% of large orders identified as potential spoofs.