Paper: arXiv 2502.18625

Authors: Jakob Albers, Mihai Cucuringu, Sam Howison, Alexander Y. Shestopaloff

Abstract

Using data from a live trading experiment on the Binance Bitcoin perpetual, we examine the effects of (i) basic order book mechanics and (ii) the persistence of price changes from immediate to short timescales, revealing the interplay between returns, queue sizes, and orders’ queue positions. We document a fundamental trade-off: a negative correlation between maker fill likelihood and post-fill returns. This dictates that viable maker strategies often require a contrarian approach, counter-trading the prevailing order book imbalance. These dynamics render commonly-cited strategies highly unprofitable, leading us to model `Reversals’: situations where a contrarian maker strategy at the touch proves effective.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 8.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper uses stochastic modeling and queueing theory (advanced math) but grounds it in a live trading experiment with real data and implementation details.

Research Flowchart

  flowchart TD
  Start["Research Goal: Uncover the trade-off between fill probability & post-fill returns for market makers using live data"] --> Data
  subgraph Data ["Data/Inputs"]
      Binance["Live Trading Experiment Data<br>(Binance BTC Perpetual)"]
  end
  Data --> Process
  subgraph Process ["Computational Analysis"]
      E1["(i) Analyze Basic Order Book Mechanics<br>(Queue Size, Positions)"]
      E2["(ii) Analyze Price Change Persistence<br>(Immediate to Short Timescales)"]
      E1 --> Int["Identify Interplay:<br>Returns vs. Queue Positions vs. Fill Likelihood"]
      E2 --> Int
  end
  Process --> Find
  subgraph Find ["Key Findings/Outcomes"]
      Trade["Documented Trade-off:<br>Negative Correlation<br>Fill Probability ↗ & Post-Fill Returns ↘"]
      Strat["Strategy Implication:<br>Viable Maker Strategies require Contrarian Approach<br>(Counter-trade order book imbalance)"]
      Rev["Model 'Reversals':<br>Situations where contrarian maker strategy at the touch is effective"]
  end

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 develops a model to identify reversals where order‑book imbalance falsely predicts the next price change and uses it for a contrarian maker strategy.

Source sentence
we develop a model to identify when the order book imbalance falsely predicts the next price change. We term this a reversal

Assumes 3

  • The analysis assumes the best possible trading fees (maker rebate -0.5 bp, taker fee 1.5 bp). execution
    Source sentence
    In the subsequent analysis, we assume the best possible trading fees; naturally, any strategy results analyzed below would only deteriorate with less favorable fees.
  • Tick size is negligible compared to fees. market
    Source sentence
    The tick size is 0.1 USD, or approximately 0.03 bp at the time of the experiment, which is negligible compared to fees.
  • Orders with estimated fill probability above 80 % are selected for modeling. model
    Source sentence
    First, we extract the subset of orders with a high estimated fill probability, specifically those exceeding a chosen threshold of 80% at submission time

Data

Backtest sample: BTC perpetual futures · crypto · intraday

  • O · Binance order-book public
    2024-08-21 to 2024-08-31 · Binance Bitcoin perpetual
    Source sentence
    This experiment ran from August 21 to 31, 2024, during which we submitted 172,800 minimum-sized maker orders

Recognised sources: Binance

Costs

Modeled Costs are part of the reported headline results.

Maker rebate -0.5 bp, taker fee 1.5 bp; best possible fees assumed in results.

Source sentence
the taker fee is 1.5 bp and the maker fee is −0.5 bp (i.e., a rebate for maker orders).

Finds 4

Reported performance (Balanced-Inventory model-based market making (Logistic Regression)): Sharpe 11.97

  • A negative correlation exists between maker fill likelihood and post‑fill returns. negative
    Source sentence
    We document a fundamental trade-off: a negative correlation between maker fill likelihood and post-fill returns.
  • Out of 172,800 submitted minimum‑sized maker orders, 95,998 filled. performance
    Source sentence
    we submitted 172,800 minimum-sized maker orders, of which 95,998 filled
  • Logistic‑Regression reversal strategy at 0.24 threshold yields average return 0.71 bp and Sharpe 11.97, vastly outperforming the naive strategy (Sharpe –109). performance
    Source sentence
    The Sharpe ratio improves from −109 for the naive strategy to 11.97 for the Logistic Regression-based strategy at the 0.24 threshold
  • The reversal‑based strategy’s return improvement is statistically significant (t‑statistic 3.062602, p‑value 0.0011). statistical
    Source sentence
    The calculated t-statistic is 3.062602 with a p-value of 0.0011, allowing us to reject the null hypothesis

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 16 related papers). Coverage is partial, so this is neither replication nor refutation.

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 (63)

  1. High Frequency Trading and The New-Market Makers (2013) — link
  2. High-Frequency Trading and Price Discovery (2014) — link
  3. Optimal Decisions in a Time Priority Queue (2018) — link
  4. Cross-impact of order flow imbalance in equity markets (2023) — link
  5. Enhancing trading strategies with order book signals (2018) — link
  6. Queue Imbalance as a One-Tick-Ahead Price Predictor in a Limit Order Book (2015) — link
  7. News Trading and Speed (2016) — link
  8. Market Making with Costly Monitoring: An Analysis of the SOES Controversy (2003)
and 55 more
  1. Toxic Arbitrage (2016) — link
  2. Statistical Predictions of Trading Strategies in Electronic Markets (2023) — link
  3. How and When are High-Frequency Stock Returns Predictable? (2022) — link
  4. The Micro-Price (2017) — link
  5. Trade arrival dynamics and quote imbalance in a limit order book (2013) — link
  6. Limit Order Strategic Placement with Adverse Selection Risk and the Role of Latency (2016) — link
  7. High Frequency Traders: Taking Advantage of Speed (2013) — link
  8. Optimal High-Frequency Market Making (2018) — link
  9. Market Making via Reinforcement Learning (2018)
  10. DeepLOB: Deep Convolutional Neural Networks for Limit Order Books (2018) — link
  11. High-frequency trading in a limit order book (2008) — link
  12. Algorithmic Trading, Stochastic Control, and Mutually Exciting Processes (2018) — link
  13. Algorithmic and High Frequency Trading (2016) — link
  14. The Negative Drift of a Limit Order Fill — → 2407.16527
  15. The Good, the Bad, and Latency: Exploratory Trading on Bybit and Binance (2024) — link
  16. Fill Probabilities in a Limit Order Book with State-Dependent Stochastic Order Flows — → 2403.02572
  17. A Model for Queue Position Valuation in a Limit Order Book (2016) — link
  18. Tackling the Problem of State Dependent Execution Probability: Empirical Evidence and Order Placement (2023) — link
  19. Deep Attentive Survival Analysis in Limit Order Books: Estimating Fill Probabilities with Convolutional-Transformers — → 2306.05479
  20. A Deep Learning Approach to Estimating Fill Probabilities in a Limit Order Book (2021) — link
  21. Towards Understanding Cryptocurrency Derivatives:A Case Study of BitMEX (2021) — link
  22. Stablecoins: Growth Potential and Impact on Banking (2022) — link
  23. Estimating covariation: Epps effect, microstructure noise (2011) — link
  24. Forecasting Volatility Using High-Frequency Data (2012) — link
  25. Backtesting (2015) — link
  26. Optimal execution with stochastic delay (2023) — link
  27. The Long Memory of Order Flow in the Foreign Exchange Spot Market (2016) — link
  28. The Role of Binance in Bitcoin Volatility Transmission (2022) — link
  29. Limit Order Books (2010) — link
  30. Fragmentation, Price Formation and Cross-Impact in Bitcoin Markets (2021) — link
  31. Latency and liquidity risk (2021)
  32. 1-share orders and trades (2017)
  33. Ultra-fast activity and intraday market quality (2019)
  34. Exploratory trading (2013)
  35. The shadow price of latency: Improving intraday fill ratios in foreign exchange markets (2021)
  36. The Case of Fleeting Orders and Flickering Quotes (2021)
  37. The Value of a Millisecond: Harnessing Information in Fast, Fragmented Markets (2016) — link
  38. Competition for Order Flow with Fast and Slow Traders (2015) — link
  39. Cross-Venue Liquidity Provision: High Frequency Trading and Ghost Liquidity (2019) — link
  40. A proposal for impact-adjusted valuation: Critical leverage and execution risk (2012)
  41. Do You Really Know Your PnL? The Importance of Impact-Adjusting the PnL (2023) — link
  42. Pseudo-Mathematics and Financial Charlatanism: The Effects of Backtest Overfitting on Out-of-Sample Performance (2014) — link
  43. The Probability of Backtest Overfitting (2012) — link
  44. Algorithmic and high-frequency trading (2015)
  45. GRASP with path-relinking for the three-index assignment problem (2000)
  46. Probability distribution of solution time in GRASP: an experimental investigation (2003)
  47. Parallel GRASP with path-relinking for job shop scheduling (1984)
  48. Title of a Book in Another Language (Quoted in that Language) (1988)
  49. Multi-start and strategic oscillation methods – principles to exploit adaptive memory (2000)
  50. Bubbles in Japan’s stock markets: a macroeconomic analysis (1995)
  51. The resurgent Japanese economy and a Japan–United States free trade agreement (1997)
  52. Efficient algorithms for layer assignment problems (1996)
  53. Gravitation (1973)
  54. Parallel GRASP with path-relinking for job shop scheduling (1983)
  55. Title of Book (2002)

3 of 63 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.