From Rules to Rewards: Reinforcement Learning for Interest Rate Adjustment in DeFi Lending

2025-05-31

AI Summary

DeFi lending faces challenges including rigid interest rate models, high default rates, and low capital efficiency. This paper proposes a dynamic interest rate governance framework based on offline reinforcement learning (RL), the first to apply TD3-BC, conservative Q-learning (CQL), and behavioral cloning to DeFi rate control. The model is trained and evaluated exclusively on historical on-chain data from Aave—including critical crisis events such as the 2021 market crash and the 2023 USDC depeg—without requiring online, on-chain training. The approach is inherently risk-aware and adaptive. Experimental results demonstrate that TD3-BC achieves the optimal trade-off among fund utilization, capital stability, and default mitigation, significantly outperforming conventional rule-based models. This validates the feasibility and effectiveness of offline RL–driven, automated, and shock-resilient interest rate governance in DeFi.

Technology Category

• Machine Learning: Reinforcement Learning
• Search and Optimization: Learning to Search
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Decentralized Finance (DeFi) lending enables permissionless borrowing via smart contracts. However, it faces challenges in optimizing interest rates, mitigating bad debt, and improving capital efficiency. Rule-based interest-rate models struggle to adapt to dynamic market conditions, leading to inefficiencies. This work applies Offline Reinforcement Learning (RL) to optimize interest rate adjustments in DeFi lending protocols. Using historical data from Aave protocol, we evaluate three RL approaches: Conservative Q-Learning (CQL), Behavior Cloning (BC), and TD3 with Behavior Cloning (TD3-BC). TD3-BC demonstrates superior performance in balancing utilization, capital stability, and risk, outperforming existing models. It adapts effectively to historical stress events like the May 2021 crash and the March 2023 USDC depeg, showcasing potential for automated, real-time governance.

Problem

• Optimizing interest rates in DeFi lending protocols
• Mitigating bad debt and improving capital efficiency
• Adapting to dynamic market conditions with RL

Innovation

• Offline Reinforcement Learning optimizes DeFi rates
• TD3-BC balances utilization, stability, and risk
• Adapts to historical stress events effectively