Allocation Mechanisms in Decentralized Exchange Markets with Frictions

A team of researchers including Mario Ghossoub, Giulio Principi, and Ruodu Wang has published a significant theoretical paper titled 'Allocation Mechanisms in Decentralized Exchange Markets with Frictions' on arXiv. The work challenges the classical economic assumption of frictionless transfers in pure-exchange economies, arguing that real-world transfers in systems like decentralized finance (DeFi) create costs. The authors propose an axiomatic study of allocation mechanisms—the rules transforming one feasible allocation of assets to another—specifically designed to account for these 'frictional costs,' which they model as a form of subadditivity in transfer expenses. This formalizes a key practical problem in blockchain-based trading.

The paper's key contribution is the axiomatic characterization of two types of mechanisms: those representable as robust (worst-case) linear allocation mechanisms and, more notably, 'Robust Conditional Mean Allocation' mechanisms, which admit representations as worst-case conditional expectations. This connects deep functional analysis and game theory (using MSC classes like 46A20 and 91B30) to the practical literature on decentralized risk-sharing pools. For developers, this provides a rigorous mathematical foundation to build and audit DEX protocols (like Uniswap or Curve) that are not just Pareto-efficient in theory but are optimized for real-world gas fees, slippage, and liquidity constraints, potentially leading to more capital-efficient decentralized markets.