Can Large Language Models Execute Parent Orders?

2026-07-30

AI Summary

This work addresses the parent order execution problem in algorithmic trading—how to split large orders into smaller child orders to minimize market impact and execution costs—by proposing the Plan-Ahead Controlled Execution (PACE) framework. PACE represents the first systematic exploration of large language models (LLMs) for this task, employing a hierarchical design that decomposes decision-making into long-term planning and short-term execution. Notably, it generalizes to new environments without requiring explicit market assumptions or task-specific training. Empirical evaluation using Level-1 market data from the Shenzhen Stock Exchange demonstrates that PACE significantly outperforms standard benchmarks including TWAP, Almgren-Chriss, and existing learning-based baselines, reducing average execution costs by 0.65 basis points. Moreover, PACE exhibits a distinctive behavioral pattern: higher confidence levels correlate with better execution performance, revealing a decision-making mechanism distinct from that of human traders.

Abstract

Parent-order execution is a core problem in algorithmic trading, where the goal is to split a large order into smaller orders while reducing execution costs. Existing approaches either rely on pre-specified market assumptions that may not hold in practice, or require task-specific training that limits adaptability to new settings. To overcome these limitations, we present the first systematic study of large language models (LLMs) for parent-order execution. This extends the use of LLMs in finance from what to trade to how to execute. We propose PACE (Plan-Ahead Controlled Execution), a hierarchical framework that decomposes parent-order execution into long-horizon planning and short-horizon execution, requiring neither explicit market assumptions nor task-specific training. Experiments on Shenzhen Stock Exchange Level-1 data show that PACE outperforms TWAP, Almgren-Chriss, and learning-based baselines, exceeding the strongest baseline by 0.65 bps. Behavioral analysis reveals that LLMs make execution decisions differently from human investors: higher model confidence predicts better performance rather than worse returns, and the model trades earlier rather than procrastinating toward the deadline. These findings suggest that LLMs can complement human traders in execution decisions.

Problem

• parent-order execution
• algorithmic trading
• execution cost
• large language models
• order splitting

Innovation

• large language models
• algorithmic trading
• parent-order execution
• PACE framework
• market microstructure