VWAP Execution with Signature-Enhanced Transformers: A Multi-Asset Learning Approach

ArXiv ID: 2503.02680 “View on arXiv”

Authors: Unknown

Abstract

In this paper I propose a novel approach to Volume Weighted Average Price (VWAP) execution that addresses two key practical challenges: the need for asset-specific model training and the capture of complex temporal dependencies. Building upon my recent work in dynamic VWAP execution arXiv:2502.18177, I demonstrate that a single neural network trained across multiple assets can achieve performance comparable to or better than traditional asset-specific models. The proposed architecture combines a transformer-based design inspired by arXiv:2406.02486 with path signatures for capturing geometric features of price-volume trajectories, as in arXiv:2406.17890. The empirical analysis, conducted on hourly cryptocurrency trading data from 80 trading pairs, shows that the globally-fitted model with signature features (GFT-Sig) achieves superior performance in both absolute and quadratic VWAP loss metrics compared to asset-specific approaches. Notably, these improvements persist for out-of-sample assets, demonstrating the model’s ability to generalize across different market conditions. The results suggest that combining global parameter sharing with signature-based feature extraction provides a scalable and robust approach to VWAP execution, offering significant practical advantages over traditional asset-specific implementations.

Keywords: VWAP Execution, Transformer, Path Signatures, Cryptocurrency, Neural Networks

Complexity vs Empirical Score

• Math Complexity: 7.5/10
• Empirical Rigor: 8.0/10
• Quadrant: Holy Grail
• Why: The paper uses advanced mathematical concepts like path signatures and transformer architectures, indicating high complexity. It also provides strong empirical evidence with backtests on 80 cryptocurrency pairs, comparing performance metrics, which suggests high implementation readiness.

  flowchart TD
    A["Research Goal:<br>Scalable VWAP Execution"] --> B["Data Collection:<br>80 Crypto Trading Pairs"]
    B --> C["Methodology:<br>GFT-Sig Transformer"]
    C --> D{"Training Strategy"}
    D --> E["Global Model<br>Multi-Asset Learning"]
    D --> F["Baseline<br>Asset-Specific Models"]
    E & F --> G["Evaluation:<br>VWAP Loss Metrics"]
    G --> H["Outcome:<br>GFT-Sig Outperforms Baselines<br>+ Cross-Asset Generalization"]