Dynamic Fee for Reducing Impermanent Loss in Decentralized Exchanges

Abstract

Decentralized exchanges (DEXs) are crucial to de-centralized finance (DeFi) as they enable trading without intermediaries. However, they face challenges like impermanent loss (IL), where liquidity providers (LPs) see their assets’ value change un-favorably within a liquidity pool compared to outside it. To tackle these issues, we propose dynamic fee mechanisms over traditional fixed-fee structures used in automated market makers (AMM). Our solution includes asymmetric fees via block-adaptive, deal-adaptive, and the "ideal but unattainable" oracle-based fee algorithm, utilizing all data available to arbitrageurs to mitigate IL. We developed a simulation-based framework to compare these fee algorithms systematically. This framework replicates trading on a DEX, considering both informed and uninformed users and a psychological relative loss factor. Results show that adaptive algorithms outperform fixed-fee baselines in reducing IL while maintaining trading activity among uninformed users. Additionally, insights from oracle-based performance underscore the potential of dynamic fee strategies to lower IL, boost LP profitability, and enhance overall market efficiency.