[Paper Review] The Limit Order Book: A Survey

Martin Gould, Mason A. Porter+4 more|arXiv (Cornell University)|Dec 1, 2010

Complex Systems and Time Series Analysis40 references8 citations

TL;DR

This survey provides a comprehensive overview of limit order books (LOBs) in financial markets, synthesizing research on their structure, dynamics, and microstructure. It reviews modeling approaches—including stochastic processes, agent-based models, and queueing theory—highlighting key findings on price formation, market impact, and resilience, offering a foundational reference for quantitative finance and market microstructure research.

ABSTRACT

Martin D. Gould, 2, ∗ Mason A. Porter, 2 Stacy Williams, Mark McDonald, Daniel J. Fenn, 2, 3 and Sam D. Howison Oxford Centre for Industrial and Applied Mathematics, Mathematical Institute, University of Oxford, Oxford OX1 3LB, UK CABDyN Complexity Centre, University of Oxford, Oxford OX1 1HP, UK FX Quantitative Research, HSBC Bank, 8 Canada Square, London E14 5HQ, UK Mathematical and Computational Finance Group, Mathematical Institute, University of Oxford, Oxford OX1 3LB, UK

Motivation & Objective

• To provide a systematic review of the theoretical and empirical literature on limit order books (LOBs) in modern financial markets.
• To identify and analyze key modeling frameworks used to understand LOB dynamics, including stochastic processes, agent-based models, and queueing theory.
• To examine the role of LOB structure in determining market microstructure phenomena such as price formation, resilience, and market impact.
• To synthesize findings on the statistical properties of order books, including the distribution of order sizes, price volatility, and the behavior of market participants.
• To serve as a foundational reference for researchers in quantitative finance, econophysics, and market microstructure by unifying diverse approaches and identifying open research questions.

Proposed method

• Systematic literature review of peer-reviewed research on limit order books from finance, econophysics, and applied mathematics.
• Categorization of modeling approaches into three main classes: stochastic processes (e.g., Hawkes processes), agent-based models, and queueing theory-based frameworks.
• Analysis of empirical data from major exchanges to validate theoretical models and extract statistical properties of order book dynamics.
• Use of mathematical formalism to describe order book state evolution, including the dynamics of bid and ask queues and the role of order arrival and cancellation processes.
• Incorporation of market impact and resilience into model frameworks, using empirical scaling laws and time-series analysis.
• Synthesis of findings across multiple asset classes and market regimes to assess model robustness and generalizability.

Experimental results

Research questions

• RQ1 How do the statistical properties of limit order books vary across different financial assets and market conditions?
• RQ2 What are the relative strengths and limitations of stochastic, agent-based, and queueing-theoretic models in capturing LOB dynamics?
• RQ3 How do order book imbalances and depth influence price formation and market resilience?
• RQ4 What is the quantitative relationship between market impact and the size and shape of the order book?
• RQ5 How do microstructure features such as order arrival rates and cancellation behavior affect the stability and efficiency of the order-driven market mechanism?

Key findings

• Limit order books exhibit scale-invariant properties, with heavy-tailed distributions for order book depth and price volatility clustering observed across multiple asset classes.
• Hawkes processes effectively model self-exciting order flow, capturing clustering of market orders and the feedback effects in price formation.
• Market impact is found to be sublinear in trade size, with the impact function scaling as √t for small trades, consistent with empirical observations.
• The resilience of the order book—the speed at which prices revert after a shock—is inversely related to market depth and order book imbalance.
• Order book microstructure is highly sensitive to the interplay between market orders, limit orders, and cancellations, with significant implications for market stability.
• Agent-based models reproduce key empirical features such as volatility clustering and autocorrelation in order flow, validating their use in simulating market behavior.