Paper: arXiv 2409.12721

Authors: Luca Lalor, Anatoliy Swishchuk

Abstract

In this paper, we study the effects of fill probabilities and adverse fills on the trading strategy simulation process. We specifically focus on a stochastic optimal control market-making problem and test the strategy on ES (E-mini S&P 500), NQ (E-mini Nasdaq 100), CL (Crude Oil) and ZN (10-Year Treasury Note), which are some of the most liquid futures contracts listed on the CME (Chicago Mercantile Exchange). We provide empirical evidence that shows how fill probabilities and adverse fills can significantly affect performance and propose a more prudent simulation framework to deal with this. Many previous works aim to measure different types of adverse selection in the limit order book (LOB), however, they often simulate price processes and market orders independently. This has the ability to largely inflate the performance of a short-term style trading strategy. Our studies show that using more realistic fill probabilities and tracking adverse fills in the strategy simulation process more accurately shows how these types of trading strategies would perform in reality.

Complexity vs Empirical Score

  • Math Complexity: 8.0/10
  • Empirical Rigor: 6.0/10 (capped from 7.0: no out-of-sample test; see penalty flags)
  • Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 7.0)

Why this score: The paper employs advanced stochastic optimal control (SOC) and probability measures for market-making, indicating high mathematical complexity, while testing on multiple real futures contracts with empirical evidence on fill probabilities and adverse fills demonstrates strong empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal: Quantify impact of fill probabilities & adverse fills on market-making strategy performance"] --> B["Data: Historical Limit Order Book Data"]
  B --> C["Methodology: Stochastic Optimal Control Market-Making Model"]
  C --> D{"Simulation Approaches"}
  D --> E["Naive Simulation<br>Independent Price/Order Matching"]
  D --> F["Prudent Simulation<br>Integrating Realistic Fill Probabilities & Adverse Fill Tracking"]
  E & F --> G["Computational Process: Backtesting on CME Futures ES, NQ, CL, ZN"]
  G --> H["Key Findings: Naive simulation inflates performance; Prudent simulation yields realistic performance metrics"]

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 enhanced simulation framework for evaluating HFT market‑making strategies that accounts for both adverse and non‑adverse trade order fills.

Source sentence
We propose an enhanced simulation framework for evaluating the performance of HFT strategies, particularly in market-making. This framework accounts for both adverse and non-adverse trade order fills in a more realistic manner

Assumes 5

  • For simplicity we assume that the fill probability ρ is constant over time (ρ∈(0,1)). model
    Source sentence
    for simplicity we assume that ρδ t = ρ with ρ ∈(0, 1)
  • Price data are resampled into 1‑second intervals using forward fill to handle irregular arrivals. data
    Source sentence
    We resampled our price data into a 1 second time interval, which was then subsequently used as the time step size in our discrete environment.
  • Each market order is assumed to have a size of 1 in the simulation. model
    Source sentence
    we currently assume that each order has a size of 1
  • Adverse fills are guaranteed to occur based on LOB system rules. execution
    Source sentence
    adverse fills are guaranteed to occur based on the rules of most LOB systems
Show 1 more
  • Non‑adverse fill probabilities are applied only to non‑adverse fills. execution
    Source sentence
    While fill probabilities are not new, they should only be applied to non‑adverse fills, as adverse fills are guaranteed to occur

Data

  • CME futures LOB data · CME order-book commercial
    April 2024 · ES, NQ, CL, ZN futures contracts
    Source sentence
    liquid U.S. futures contracts, calculated using LOB data from April 9, 2024

Costs

Unclear The classifier was not confident.

Finds 3

  • The benchmark simulation environment inflates performance; the improved environment shows much worse results. performance
    Source sentence
    the strategy performed reasonably well in the benchmark environment and much worse in our improved environment
  • Excluding adverse fills and using unrealistic non‑adverse fill probabilities can significantly inflate strategy performance. performance
    Source sentence
    excluding adverse fills and non‑adverse fill probabilities can significantly inflate the performance of a trading strategy
  • Lower non‑adverse fill probabilities and the presence of adverse fills substantially affect wealth and inventory paths. performance
    Source sentence
    have significantly affected how these processes evolved

Checked by others

Supported by

  • The Negative Drift of a Limit Order Fill (2024)
    On the question: How does the assumed fill probability impact observed strategy performance?
    Compare the findings

    This paper: Excluding adverse fills and using unrealistic non‑adverse fill probabilities can significantly inflate strategy performance.

    That paper: The empirical fill probability is much lower than the assumed 100% in many prior models.

    Both note that using unrealistic 100% fill rates (A) inflates performance; B shows excluding adverse fills (non‑adverse probabilities) leads to over‑optimistic results, confirming A's lower empirical fill probability.

Cited by 5 papers in the archive (2 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 (25)

  1. Deep attentive survival analysis in limit order books: Estimating fill probabilities with convolutional-transformers — → 2306.05479
  2. An empirical analysis of the limit order book and the order flow in the Paris Bourse (1995)
  3. Price discovery without trading: Evidence from limit orders (2019)
  4. Optimal execution of limit and market orders with trade director, speed limiter, and fill uncertainty (2017)
  5. Enhancing trading strategies with order book signals (2018)
  6. Algorithmic and high-frequency trading (2015)
  7. Buy low, sell high: A high frequency trading perspective (2014)
  8. Algorithmic trading, stochastic control, and mutually exciting processes (2018)
and 17 more
  1. The price impact of order book events (2014)
  2. Price dynamics in a Markovian limit order market (2013)
  3. The negative drift of a limit order fill — → 2407.16527
  4. Flow toxicity and liquidity in a high-frequency world (2012)
  5. Limit order books (2013)
  6. Low-latency trading (2013)
  7. A dynamic limit order market with fast and slow traders (2014)
  8. Market Making at the Touch with Short-Term Alpha (Chap 10.4.2) (2019)
  9. Limit Order Book Simulations: A Review — → 2402.17359
  10. Market making under a weakly consistent limit order book model (2019)
  11. Fill Probabilities in a Limit Order Book with State-Dependent Stochastic Order Flows — → 2403.02572
  12. A deep learning approach to estimating fill probabilities in a limit order book (2022)
  13. Optimal Liquidation, Acquisition and Market Making Problems in HFT under Hawkes Models for LOB (2022)
  14. Multi-period portfolio optimisation with alpha decay (2018)
  15. A semi- Markovian modeling of limit order markets (2017)
  16. Compound Hawkes processes in limit order books (2019)
  17. The alpha life cycle of quantitative strategy (2017)

4 of 25 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.