Learning Market Making with Closing Auctions

2026/01/31 by Julius Graf, Thibaut Mastrolia
Economics, Econometrics and Finance · Mathematics · #q-fin.TR #math.OC

paper · pdf

arxiv created 2026/07/30 · arxiv updated 2026/07/31

Abstract

In this work, we investigate a market making execution problem on a trading session in which a continuous phase on a limit order book is followed by a closing auction. Whereas standard optimal market making models typically rely on terminal inventory penalties to manage end-of-day risk, ignoring the significant liquidity events available in closing auctions, we propose a deep reinforcement learning framework, consisting of a Deep Q-Network and its continuous-control actor-critic extensions (DDPG, TD3 and SAC), that explicitly incorporates this mechanism. We introduce a market making framework designed to explicitly anticipate the closing auction, continuously refining the projected clearing price as the trading session evolves. We develop a generative stochastic market model to simulate the trading session and to emulate the market. Our theoretical model and these deep reinforcement learning methods are applied on the generator in two settings: (1) when the mid price follows a rough Heston model with generative data from this stochastic model; and (2) when the mid price corresponds to historical data of assets from the S&P 500 index and the performance of our algorithm is compared with stylized reference benchmarks from optimal market making.

Citations

• Optimal Rebate Design: Incentives, Competition and Efficiency in Auction Markets
• Market Making without Regret
• Logarithmic regret in the ergodic Avellaneda-Stoikov market making model
• Clearing time randomization and transaction fees for auction market design
• Equilibria and incentives for illiquid auction markets
• Market Making with Scaled Beta Policies
• On the discrete-time simulation of the rough Heston model
• AHEAD : Ad-Hoc Electronic Auction Design
• Learning a functional control for high-frequency finance
• Reinforcement Learning for Market Making in a Multi-agent Dealer Market
• Deep reinforcement learning for market making in corporate bonds:\n beating the curse of dimensionality
• Optimal auction duration: A price formation viewpoint
• A Theoretical Analysis of Deep Q-Learning
• Double Deep Q-Learning for Optimal Execution
• Soft Actor-Critic Algorithms and Applications
• Portfolio Optimization in Fractional and Rough Heston Models
• Market Making via Reinforcement Learning
• Addressing Function Approximation Error in Actor-Critic Methods
• Multi-factor approximation of rough volatility models
• Deep Learning for Limit Order Books
• Deep Reinforcement Learning with Double Q-learning
• Continuous control with deep reinforcement learning
• Volatility is rough
• Playing Atari with Deep Reinforcement Learning

Related

• Optimal Market Making by Reinforcement Learning
• Market Making Strategies with Reinforcement Learning
• Market Making via Reinforcement Learning
• Market Making with Deep Reinforcement Learning from Limit Order Books
• Model-based Reinforcement Learning for Predictions and Control for Limit Order Books