Reproduce “Market Making with Scaled Beta Policies” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Joseph Jerome, Gregory Palmer, Rahul Savani
A new parametric representation for automated market maker actions, using scaled beta distributions, enables flexible volume placement across price levels. This approach demonstrates effective inventory management solely through limit orders, avoiding costly market orders, and achieves positive mean returns in high-fidelity simulations.
Key insights
- Scaled beta distributions offer a continuous, highly flexible, and unifying parametric action space for automated market makers, generalizing prior discrete and rigid approaches.
- The ability to dynamically skew volume profiles using scaled beta distributions allows for sophisticated inventory management solely through limit order placement, avoiding costly market orders.
- The continuous and low-dimensional parameterization of scaled beta policies provides a suitable action space for training advanced reinforcement learning agents.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.