Paper: arXiv 2410.08744

Authors: Konark Jain, Jean-François Muzy, Jonathan Kochems, Emmanuel Bacry

Abstract

Tick-sizes not only influence the granularity of the price formation process but also affect market agents’ behavior. We investigate the disparity in the microstructural properties of the Limit Order Book (LOB) across a basket of assets with different relative tick-sizes. A key contribution of this study is the identification of several stylized facts, which are used to differentiate between large, medium, and small-tick assets, along with clear metrics for their measurement. We provide cross-asset visualizations to illustrate how these attributes vary with relative tick-size. Further, we propose a Hawkes Process model that {"\color{black"}not only fits well for large-tick assets, but also accounts for }sparsity, multi-tick level price moves, and the shape of the LOB in small-tick assets. Through simulation studies, we demonstrate the {"\color{black"} versatility} of the model and identify key variables that determine whether a simulated LOB resembles a large-tick or small-tick asset. Our tests show that stylized facts like sparsity, shape, and relative returns distribution can be smoothly transitioned from a large-tick to a small-tick asset using our model. We test this model’s assumptions, showcase its challenges and propose questions for further directions in this area of research.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper heavily employs advanced stochastic modeling (Hawkes processes, SPDEs, generator operators) with complex mathematical formulations, while also providing empirical analysis of stylized facts, cross-asset visualizations, and simulation studies to test model versatility.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Modeling LOB dynamics across<br>different tick-size regimes"] --> B["Data Collection<br>Empirical LOB data from<br>equity markets with varying<br>relative tick-sizes"]
  B --> C["Methodology<br>Hawkes Process modeling<br>to capture event clustering"]
  C --> D{"Computational Analysis<br>Simulation & Parameter Estimation"}
  D --> E["Key Findings<br>Identified stylized facts<br>differentiating tick regimes"]
  E --> F["Outcomes<br>General model valid for both<br>large & small tick assets<br>Smooth transition between regimes"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes an extension of the Compound Hawkes Model, including meta‑queues, to model both large‑tick and small‑tick limit order books.

Source sentence
(ii) In Section 3, we propose an extension to the Compound Hawkes Model in Jain et al. 2024a which is also capable of modeling small-tick assets with a sparse LOB and wide spreads.

Assumes 4

  • All assets have a constant tick size of δ = 1 cent. market
    Source sentence
    Let us point out that the assets in our collection are all US equities with large capitalization and therefore have all a constant tick size δ = 1 cent 2.
  • Relative tick‑size (ε) defined as δ divided by average mid‑price (in bps) is used as the proxy for comparing tick sizes across assets. model
    Source sentence
    5 Thus, in our study, as a proxy for comparing the tick size of the various assets, we prefer to use the relative tick-size (measured in basis points, bps) defined by ϵ := δ ⟨Pmid⟩× 104 (3) As shown in Table 1, the relative tick size of the assets of our collections covers a very wide range of values from 16.47 bps (for SIRI) to 0.05 bps (for BKNG).
  • Order sizes, order arrival distances and unseen queue sizes are assumed i.i.d. random variables. statistical
    Source sentence
    Assumption 2 (i) Order sizes, (ii) Order arrival distances and (iii) Unseen Queue sizes on queue depletion, are i.i.d. random variables.
  • Limit orders (LOs) and cancel orders (COs) are assumed to be independent of intensity functions except for the sparsity variable. model
    Source sentence
    Therefore we conclude that though for LOs and COs, we can assume the LOB state variables except QT to be independent of their intensity functions, there exists a clear dependence of in-tensities of in-spread orders on the current mT , mD.

Data

  • LOBSTER tick‑by‑tick order‑book data · LOBSTER order-book commercial
    2nd January 2019 to 30th December 2019 · 15 US large‑cap equities (S&P 500 constituents)
    Source sentence
    For each asset, high-resolution, tick-by-tick LOB data obtained from the LOBSTER provider (LOBSTER: Limit Order Book System - The Efficient Reconstructor n.d.) is employed. m-tick PM 2.61 81.89 1.22 Medium-tick MSFT 1.25 129.44 0.77 Medium-tick IBM 3.37 137.94 0.72 Medium-tick AAPL 1.97 208.62 0.48 Medium-tick TSLA 12.66 270.94 0.36 Small-tick CHTR 23.76 394.76 0.25 Small-tick GOOG 57.98 1186.57 0.08 Small-tick AMZN

Recognised sources: LOBSTER

Costs

Not applicable The paper reports no trading or execution result.

Finds 3

  • The average bid‑ask spread is highly correlated with relative tick‑size, showing a constant 1‑tick spread for large‑tick assets and a power‑law decay (exponent ≈ ‑1.23) for small‑ and medium‑tick assets. statistical
    Source sentence
    Bid-Ask Spread : Mean and Distribution vs Relative Tick-Size Figure 1.: Average Spread (in ticks) vs Relative Tick Size(in bps) in a log-log scale : Remarkably, the average spread appears to be highly correlated with the relative tick size. et al. 2018) the spread seems to be a power-law (Eq. (4)) of the relative tick-size with an exponent close to -1 (the slope of the orange plain line is -1.23).
  • Simulation of the proposed Hawkes model reproduces the empirical scaling exponents for index of dispersion, average mid‑price change, and LOB shape peak across tick‑size regimes. theoretical
    Source sentence
    This demonstrates that the MQH model is able to replicate not only 28 Metric Symbolic Form Simulated Slope Empirical Slope (from Section 2) Index of Dispersion D ∼ϵαD αsim D = −1.43 αemp D = −1.36 Mean Mid-Price Change ⟨rmid⟩∼ϵαr αsim r = −1.05 αemp r = −0.95 Maxima of Avg. In Section 2 we established three power-law relationships between relative tick size and (i) the index of dispersion of the spread, (ii) the
  • Two distinct regimes in spread behavior are identified: a constant‑spread regime for assets with ε ≥ 1.5 and a regime where spread follows a power‑law for ε < 1.5. statistical figure from elsewhere in the paper
    Source sentence
    Bid-Ask Spread : Mean and Distribution vs Relative Tick-Size Figure 1.: Average Spread (in ticks) vs Relative Tick Size(in bps) in a log-log scale : Remarkably, the average spread appears to be highly correlated with the relative tick size. The figure clearly reveals two regimes.

Checked by others

Supported by

  • Stylized Facts and Market Microstructure: An In-Depth Exploration of German Bond Futures Market (2024)
    On the question: How does tick size affect the bid‑ask spread behavior in bond futures markets?
    Compare the findings

    This paper: The average bid‑ask spread is highly correlated with relative tick‑size, showing a constant 1‑tick spread for large‑tick assets and a power‑law decay (exponent ≈ ‑1.23) for small‑ and medium‑tick assets.

    That paper: For assets with large tick sizes, the bid‑ask spread distribution is limited to a few possible values.

    A finds large‑tick assets have spreads limited to few discrete values; B shows large‑tick assets keep a constant 1‑tick spread and small/medium ticks follow a power‑law decay with exponent ≈‑1.23.

Cited by 3 papers in the archive (1 by independent authors)

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (135)

  1. A nonparametric estimation procedure for the Hawkes process: comparison with maximum likelihood estimation (2018) — link
  2. A Stochastic Model for Order Book Dynamics (2010)
  3. Statistical theory of the continuous double auction (2003) — link
  4. The predictive power of zero intelligence in financial markets (2005) — link
  5. Statistical properties of stock order books: empirical results and models (2002) — link
  6. Limit order books (2013) — link
  7. The Price Impact of Order Book Events (2013) — link
  8. Price Dynamics in a Markovian Limit Order Market (2013) — link
and 127 more
  1. DeepLOB : Deep Convolutional Neural Networks for Limit Order Books (2019) — link
  2. Quant GANs : deep generation of financial time series (2020) — link
  3. Universal features of price formation in financial markets: perspectives from Deep Learning (2018) — link
  4. A steady-state model of the continuous double auction (2003) — link
  5. Statistical Modeling of High-Frequency Financial Data (2011) — link
  6. Simulating and Analyzing Order Book Data: The Queue-Reactive Model (2015)
  7. Modelling high-frequency limit order book dynamics with support vector machines (2015) — link
  8. Modeling financial time-series with generative adversarial networks (2019) — link
  9. A Mathematical Approach to Order Book Modelling (2011) — link
  10. An agent based model of the E-Mini S&P 500 applied to flash crash analysis (2012) — link
  11. Order book dynamics in liquid markets: limit theorems and diffusion approximations (2011) — link
  12. Sig-Wasserstein GANs for Time Series Generation (2022) — link
  13. Hawkes Processes in Finance (2015) — link
  14. Ergodicity and scaling limit of a constrained multivariate Hawkes process (2014) — link
  15. Estimation of slowly decreasing hawkes kernels: application to high-frequency order book dynamics (2016)
  16. " Market making" behaviour in an order book model and its impact on the bid-ask spread (2010) — link
  17. Measuring the resiliency of an electronic limit order book (2007) — link
  18. Long-Time Behavior of a Hawkes Process–Based Limit Order Book (2015) — link
  19. A buffer Hawkes process for limit order books (2017) — link
  20. High-dimensional Hawkes processes for limit order books: modelling, empirical analysis and numerical calibration (2018) — link
  21. From asymptotic properties of general point processes to the ranking of financial agents (2019) — link
  22. A scaling limit for limit order books driven by Hawkes processes (2019)
  23. Hawkes processes and their applications to finance: a review (2018) — link
  24. Deep Hawkes process for high-frequency market making (2021) — link
  25. Abides: Towards high-fidelity market simulation for ai research (2019) — link
  26. Generating Realistic Stock Market Order Streams (2020) — link
  27. High frequency asymptotics for the limit order book (2016)
  28. State-dependent Hawkes processes and their application to limit order book modelling (2022)
  29. A law of large numbers for limit order books (2017)
  30. Modeling high-frequency order flow imbalance by functional limit theorems for two-sided risk processes (2015)
  31. A weak law of large numbers for a limit order book model with fully state dependent order dynamics (2017)
  32. Hydrodynamic limit of order-book dynamics (2018)
  33. A Markov model of a limit order book: thresholds, recurrence, and trading strategies (2018)
  34. Ergodicity and diffusivity of Markovian order book models: a general framework (2017)
  35. Deep learning modeling of limit order book: A comparative perspective (2020) — link
  36. Estimation of an Order Book Dependent Hawkes Process for Large Datasets — → 2307.09077
  37. A one-level limit order book model with memory and variable spread (2017)
  38. Algorithmic trading with Markov chains (2010)
  39. Limit order books, diffusion approximations and reflected SPDEs: from microscopic to macroscopic models (2020)
  40. Second order approximations for limit order books (2018)
  41. Second-Order Approximation of Limit Order Books in a Single-Scale Regime — → 2308.00805
  42. Order-book modeling and market making strategies (2018)
  43. Modeling Bid and Ask Price Dynamics with an Extended Hawkes Process and Its Empirical Applications for High-Frequency Stock Market Data (2022)
  44. Hawkes-based models for high frequency financial data (2022)
  45. A stochastic partial differential equation model for limit order book dynamics (2021)
  46. Dynamic equilibrium limit order book model and optimal execution problem (2014) — link
  47. Equilibrium Model of Limit Order Books: A Mean-Field Game View (2022)
  48. Order book dynamics with liquidity fluctuations: limit theorems and large deviations (2020) — link
  49. Dynamic calibration of order flow models with generative adversarial networks (2022)
  50. Limit Order Book Simulation with Generative Adversarial Networks (2023)
  51. Intra-Day Price Simulation with Generative Adversarial Modelling of the Order Flow (2021)
  52. Generative adversarial nets (2014)
  53. Conditional generative adversarial nets (2014) — link
  54. Wasserstein generative adversarial networks (2017)
  55. The price impact of order book events (2014)
  56. Price discovery and learning during the preopening period in the Paris Bourse (1999)
  57. Limit order book as a market for liquidity (2005)
  58. The lob recreation model: Predicting the limit order book from taq history using an ordinary differential equation recurrent neural network (2021)
  59. The Limit Order Book Recreation Model (LOBRM): An Extended Analysis (2021)
  60. State Dependent Parallel Neural Hawkes Process for Limit Order Book Event Stream Prediction and Simulation (2022)
  61. Neural Stochastic Agent-Based Limit Order Book Simulation: A Hybrid Methodology — → 2303.00080
  62. Fast agent-based simulation framework of limit order books with applications to pro-rata markets and the study of latency effects (2020) — link
  63. Deep learning for market by order data (2021)
  64. Generative AI for End-to-End Limit Order Book Modelling: A Token-Level Autoregressive Generative Model of Message Flow Using a Deep State Space Network — → 2309.00638
  65. Marked Hawkes process modeling of price dynamics and volatility estimation (2017)
  66. The neural hawkes process: A neurally self-modulating multivariate point process (2017)
  67. Deep Learning (2016)
  68. Machine Learning and Data Sciences for Financial Markets: A Guide to Contemporary Practices (2023)
  69. Probability distribution of returns in the Heston model with stochastic volatility* (2002) — link
  70. Learning to simulate realistic limit order book markets from data as a World Agent (2022)
  71. Conditional Generators for Limit Order Book Environments: Explainability, Challenges, and Robustness — → 2306.12806
  72. Limit Order Books (2016)
  73. Hawkes Process: Fast Calibration, Application to Trade Clustering, and Diffusive Limit (2014) — link
  74. Hawkes model specification for limit order books (2022)
  75. High-frequency simulations of an order book: a two-scale approach (2011)
  76. Econophysics review: II. Agent-based models (2011)
  77. Get real: Realism metrics for robust limit order book market simulations (2020)
  78. The probability of backtest overfitting (2016)
  79. Pseudomathematics and financial charlatanism: The effects of backtest over fitting on out-of-sample performance (2014)
  80. A data-driven market simulator for small data environments (2020) — link
  81. An estimation procedure for the Hawkes process (2017)
  82. Queue-reactive Hawkes models for the order flow (2019) — link
  83. The role of volume in order book dynamics: a multivariate Hawkes process analysis (2017)
  84. Tick: a Python library for statistical learning, with a particular emphasis on time-dependent modelling (2017) — link
  85. Market impacts and the life cycle of investors orders (2015)
  86. Order Flow and Price Formation (2023)
  87. Deciphering How Investors’ Daily Flows are Forming Prices (2023)
  88. Data-snooping, technical trading rule performance, and the bootstrap (1999)
  89. A reality check for data snooping (2000)
  90. Uncovering causality from multivariate Hawkes integrated cumulants (2018)
  91. Reinforcement Learning in Non - Markovian Environments (2024) — link
  92. Value Function Based Reinforcement Learning in Changing Markovian Environments
  93. Simple Agent , Complex Environment : Efficient Reinforcement Learning with Agent States
  94. Non- Markovian Reinforcement Learning using Fractional Dynamics (2021) — link
  95. Q- Learning for Stochastic Control under General Information Structures and Non - Markovian Environments (2024) — link
  96. Labeling Q -learning for non- Markovian environments (1999) — link
  97. Policy Gradient Algorithms with Monte - Carlo Tree Search for Non - Markov Decision Processes (2022) — link
  98. Approximate information state based convergence analysis of recurrent Q -learning (2023) — link
  99. On Q -learning Convergence for Non - Markov Decision Processes (2018) — link
  100. High dimensional Hawkes processes (2014) — link
  101. Reinforcement Learning , Bit by Bit (2023) — link
  102. Trades, quotes and prices: financial markets under the microscope (2018)
  103. Tick Size and Price Diffusion (2011) — link
  104. Deep Limit Order Book Forecasting — → 2403.09267
  105. HLOB–Information Persistence and Structure in Limit Order Books — → 2405.18938
  106. Limit Order Book Simulations: A Review — → 2402.17359
  107. Limit Order Book dynamics and order size modelling using Compound Hawkes Process — → 2312.08927
  108. LOBSTER: Limit Order Book System - The Efficient Reconstructor
  109. Maximum Likelihood Estimation in the Weibull Distribution Based on Complete and on Censored Samples (1965)
  110. Optimal Tick Size (2023) — link
  111. Relative Tick Size and the Trading Environment (2018) — link
  112. When Bigger is Better: The Impact of a Tiny Tick Size on Undercutting Behavior (2023)
  113. Market Making, the Tick Size, and Payment-for-Order Flow: Theory and Evidence (1995)
  114. Nasdaq market structure and spread patterns (1997) — link
  115. Tick Size Pilot Program and price discovery in U.S. stock markets (2022) — link
  116. Why Discrete Price Fragments U.S. Stock Exchanges and Disperses Their Fee Structures (2018) — link
  117. Tick size, liquidity for small and large orders, and price informativeness: Evidence from the Tick Size Pilot Program (2020) — link
  118. From Glosten-Milgrom to the whole limit order book and applications to financial regulation (2019) — link
  119. A New Approach for the Dynamics of Ultra-High-Frequency Data: The Model with Uncertainty Zones (2010) — link
  120. Empirical properties of asset returns: stylized facts and statistical issues (2001) — link
  121. What is the optimal tick size? a cross-sectional analysis of execution costs on NASDAQ (2017)
  122. Large Tick Assets: Implicit Spread and Optimal Tick Size (2015) — link
  123. On Bid and Ask Side-Specific Tick Sizes (2023) — link
  124. The long memory of the efficient market (2004)
  125. How markets slowly digest changes in supply and demand (2009)
  126. Order-splitting and long-memory in an order-driven market (2010)
  127. Why is equity order flow so persistent? (2015)

9 of 135 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.