An adaptive dual-level reinforcement learning approach for optimal trade execution

Soohan Kim, Jimyeong Kim, Hong Kee Sul, Youngjoon Hong

Abstract: Highlights•We propose a novel reinforcement learning approach for optimal trade execution.•We propose the U-shape Transformer model to capture the day-to-day oscillation.•Ablation studies show the strengths of our methods in approximating the VWAP.