Price Discovery on Decentralized Exchanges

Abstract

In contrast to centralized exchanges, decentralized exchanges (DEXs) process orders in discrete time and require traders to bid a priority fee to determine the execution priority. We employ a structural vector-autoregressive model to provide evidence that the priority fee reveals the private information of a DEX trade, contributing to price discovery. A one standard deviation shock in the high-fee DEX trade flow leads to a permanent price impact between 4.27 and 8.16 basis points. We show that informed traders bid high fees not only to reduce execution risk but also to compete with each other. Using a unique dataset of Ethereum mempool orders, we lend support to the hypothesis that informed traders primarily compete on DEXs following a jump-bidding strategy.

Keywords: price discovery, blockchain fees, decentralized exchanges, informed traders, private information

JEL Classification: G10, G23, G14