Empirical Analysis of High-Frequency Algorithmic Trading Activity, Market Liquidity, and Short-Term Price Volatility in Uniswap V3 Pools

Abstract

High-frequency trading (HFT) algorithms execute millions of transactions daily, yet their net impact on decentralized finance (DeFi) market quality remains poorly understood. While traditional financial theory suggests that increased HFT activity can enhance liquidity dimension of market quality and influence price volatility, comparable empirical evidence for decentralized exchanges remains limited. This study analyzes hourly data from eleven major Uniswap v3 pools on Ethereum (June 2025 – December 2025) to examine the relationship between algorithmic arbitrage activity, liquidity, and short-term volatility. Algorithmic trading activity is proxied through atomic arbitrage transactions, from which two measures are constructed: arbitrage share and arbitrage volume share. Market quality is assessed using realized volatility and the Amihud illiquidity ratio. Two-way fixed-effects panel regressions, controlling for pool-specific heterogeneity, market shocks, and trading volume, reveal no significant relationship between arbitrage activity and short-term volatility (p > 0.12 for both measures). However, arbitrage volume share negatively correlates with illiquidity (beta = -0.0028, p = 0.0196), suggesting improved liquidity. These findings suggest algorithmic arbitrage enhances liquidity without increasing price instability in Ethereum-based DeFi markets, offering empirical support for the positive role of automated trading in decentralized financial markets.