RAmmStein: Regime Adaptation in Mean-reverting Markets with Stein Thresholds — Optimal Impulse Control in Concentrated AMMs
Abstract
We address the problem of optimal liquidity management in concentrated automated market makers (AMMs) using optimal control theory. RAmmStein formulates the rebalancing decision as a Hamilton-Jacobi-Bellman quasi-variational inequality and solves it via deep reinforcement learning, incorporating mean-reversion dynamics to learn when to rebalance. Evaluated on 6.8M Coinbase trades, the approach achieves 1.60% net ROI while reducing rebalancing frequency by 85% compared to greedy strategies.
Type
Publication
arXiv preprint
Authors
Senior Research Scientist
Pranay Anchuri is a Senior Research Scientist at Offchain Labs. His research spans
blockchain protocols, verifiable computation, and machine learning applied to
decentralized systems. He has published at top venues including KDD, JMLR, and ICDM,
and is an inventor on seven US patents. He holds a PhD in Computer
Science from Rensselaer Polytechnic Institute.