Thinking Fast and Slow: Data-Driven Adaptive DeFi Borrow-Lending Protocol

TL;DR

This work tackles instability in DeFi borrow-lending caused by collateral volatility by introducing a two-tier, data-driven protocol. A fast online interest-rate controller uses a least-squares estimator to learn and promptly converge to the equilibrium rate, while a slow collateral-factor planner adapts long-horizon risk and utilization via robust optimization. The approach comes with theoretical guarantees on convergence and adversarial robustness, plus empirical validation showing faster convergence and stable utilization compared with fixed-curve baselines. The protocol promises more stable liquidity, fairer competition with external markets, and practical feasibility through blockchain-friendly, rollup-based implementation.

Abstract

Decentralized finance (DeFi) borrowing and lending platforms are crucial to the decentralized economy, involving two main participants: lenders who provide assets for interest and borrowers who offer collateral exceeding their debt and pay interest. Collateral volatility necessitates over-collateralization to protect lenders and ensure competitive returns. Traditional DeFi platforms use a fixed interest rate curve based on the utilization rate (the fraction of available assets borrowed) and determine over-collateralization offline through simulations to manage risk. This method doesn't adapt well to dynamic market changes, such as price fluctuations and evolving user needs, often resulting in losses for lenders or borrowers. In this paper, we introduce an adaptive, data-driven protocol for DeFi borrowing and lending. Our approach includes a high-frequency controller that dynamically adjusts interest rates to maintain market stability and competitiveness with external markets. Unlike traditional protocols, which rely on user reactions and often adjust slowly, our controller uses a learning-based algorithm to quickly find optimal interest rates, reducing the opportunity cost for users during periods of misalignment with external rates. Additionally, we use a low-frequency planner that analyzes user behavior to set an optimal over-collateralization ratio, balancing risk reduction with profit maximization over the long term. This dual approach is essential for adaptive markets: the short-term component maintains market stability, preventing exploitation, while the long-term planner optimizes market parameters to enhance profitability and reduce risks. We provide theoretical guarantees on the convergence rates and adversarial robustness of the short-term component and the long-term effectiveness of our protocol. Empirical validation confirms our protocol's theoretical benefits.

Figures (3)

• Figure 1: Protocol Overview. The interest rate controller observes borrower actions to estimate $r^*$ and set $r_t\xspace = \hat{r}^*$. The collateral factor planner includes a parameter estimator and an optimizer: the estimator finds $r_{o}^l\xspace$, $r_{o}^b\xspace$, and $\sigma$, while the optimizer uses these estimates to determine the optimal collateral factor for the market.
• Figure 2: Comparing the LSE-based interest rate controller and the piecewise linear curve.
• Figure 3:

Theorems & Definitions (12)

• Lemma 1: Maximum loan-to-value adoption
• Definition 2: Market equilibrium
• Lemma 3: Simplified default and price change terms
• Theorem 4
• Definition 5: Rate of equilibrium convergence
• Definition 6: Optimality Index
• Definition 7: Adversarial Susceptibility
• Theorem 8: LSE convergence rate
• Theorem 9: Baseline convergence rate
• Lemma 10