Citation

Ladóczki, B., Rásonyi, M., Tapolcai, J. “Where Does MEV Really Come From? Revisiting CEX–DEX Arbitrage on Ethereum.” arXiv:2604.15973v1 [cs.CR] (17 Apr 2026). BME / HUN-REN Rényi Institute / ELTE, Budapest.

Core Question

Where does MEV revenue originate, and how “dark” is it (i.e. does it harm users)? CEX–DEX arbitrage is the benign form — it keeps AMM prices aligned with the external market. But theory has underestimated its scale while empirics show it dwarfs atomic arb and sandwiches by orders of magnitude. The paper reconciles the gap.

Key Argument

  • Most prior AMM models use the Black–Scholes SDE (geometric Brownian motion) with continuous price paths — prices move only in small increments.
  • BS underestimates arbitrage profit by ignoring price jumps, which are precisely where arbitrage opportunities arise.
  • They build an extended discrete-time AMM model where the price process = diffusive component + stochastic jumps with arbitrary noise distributions.
  • A general discrete-time SDE lets them compute the stationary probability distribution via function iteration with geometric convergence, and they prove the mispricing process is an ergodic Markov chain.
  • Implemented in C++; fit to Ethereum spot + AMM data (Uniswap V2 ETH–USDT pool).

Headline Findings

  • Corrected theory shows CEX–DEX arbitrage requires trading volumes on the order of the total activity of major liquidity pools and yields profits comparable to total MEV — matching empirical observations that earlier (BS-based) models missed by orders of magnitude.
  • Provides a natural theoretical explanation for why CEX–DEX volume exceeds atomic arb and sandwich volume by several orders of magnitude.

Connection to Wiki

See Also