Paper: arXiv 2601.22113

Authors: Robert de Witt, Mikko S. Pakkanen

Abstract

We present the first application of MAP-Elites, a quality-diversity algorithm, to trade execution. Rather than searching for a single optimal policy, MAP-Elites generates a diverse portfolio of regime-specialist strategies indexed by liquidity and volatility conditions. Individual specialists achieve 8-10% performance improvements within their behavioural niches, while other cells show degradation, suggesting opportunities for ensemble approaches that combine improved specialists with the baseline PPO policy. Results indicate that quality-diversity methods offer promise for regime-adaptive execution, though substantial computational resources per behavioural cell may be required for robust specialist development across all market conditions. To ensure experimental integrity, we develop a calibrated Gymnasium environment focused on order scheduling rather than tactical placement decisions. The simulator features a transient impact model with exponential decay and square-root volume scaling, fit to 400+ U.S. equities with $R^2>0.02$ out-of-sample. Within this environment, two Proximal Policy Optimization architectures - both MLP and CNN feature extractors - demonstrate substantial improvements over industry baselines, with the CNN variant achieving 2.13 bps arrival slippage versus 5.23 bps for VWAP on 4,900 out-of-sample orders ($21B notional). These results validate both the simulation realism and provide strong single-policy baselines for quality-diversity methods.

Complexity vs Empirical Score

  • Math Complexity: 5.5/10
  • Empirical Rigor: 8.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: This paper presents an innovative application of quality-diversity algorithms (MAP-Elites) to optimal trade execution, backed by a highly rigorous empirical setup. The authors calibrate a transient impact propagator model on 400+ U.S. equities and evaluate RL agents on a massive out-of-sample dataset ($21B notional), demonstrating significant slippage reduction over standard benchmarks.

Research Flowchart

  flowchart TD
    A[Goal: Develop Regime-Adaptive Trade Execution Strategies] --> B[Data: 400+ US Equities & 4,900 OOS Orders]
    B --> C[Environment: Calibrated Gymnasium Simulator with Transient Impact Model]
    C --> D[Baseline Optimization: Train MLP & CNN PPO Policies]
    D --> E[Quality-Diversity: Apply MAP-Elites across Liquidity & Volatility Niches]
    E --> F[Findings: CNN PPO beats VWAP 2.13 vs 5.23 bps; MAP-Elites Specialists gain 8-10% in-niche]