Scholarlink

This paper investigates the application of deep reinforcement learning (DRL) to optimize market-making strategies within limit order book environments. We introduce a DRL-based market-making agent that leverages structural features of limit order book data to inform dynamic trading decisions. The agent’s reward function is explicitly formulated around two critical factors: market liquidity and inventory risk, thereby establishing a novel decision-making framework for complex financial settings. Experimental results demonstrate that the agent exhibits strong performance in capturing bid-ask spreads and effectively managing inventory exposure. These findings offer valuable theoretical insights and practical implications for the design of market-making strategies in high-frequency trading.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep Reinforcement Learning-Based Market Making in Limit Order Books

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Shaochangyi Luo

摘要

This paper investigates the application of deep reinforcement learning (DRL) to optimize market-making strategies within limit order book environments. We introduce a DRL-based market-making agent that leverages structural features of limit order book data to inform dynamic trading decisions. The agent’s reward function is explicitly formulated around two critical factors: market liquidity and inventory risk, thereby establishing a novel decision-making framework for complex financial settings. Experimental results demonstrate that the agent exhibits strong performance in capturing bid-ask spreads and effectively managing inventory exposure. These findings offer valuable theoretical insights and practical implications for the design of market-making strategies in high-frequency trading.