Optimal Execution in a Queue Reactive Model
Presentation
@ University of Pavia — Fintech Research Conference
I presented at the Fintech Research Conference at the University of Pavia on March 27, 2026, on Reinforcement Learning for Optimal Execution with the Queue Reactive Model.
My presentation covered the formulation of optimal execution as a reinforcement learning problem within the queue reactive model framework. I discussed the advantages of the queue reactive model for market simulation compared to stochastic and agent-based alternatives, how to embed the execution problem into an RL environment, and presented experimental results showing how RL agents learn to minimize market impact and transaction costs in a realistic order book setting.