Imperfect Commitment in Maximal Extractable Value Auctions

Battle Count: 51

The paper is highly relevant to MEV extraction strategies, which constitute a specialized form of quantitative trading in decentralized finance. It directly informs searcher bidding strategies (risky vs. deterrence bids), the economics of sandwich attacks, naked arbitrage, liquidations, and backruns. The type-specific heterogeneity in replicability γ(τ) and competitive intensity has direct implications for MEV strategy selection and pricing. However, it is more focused on mechanism design and auction theory than on traditional quantitative trading signals or portfolio construction. The findings on builder defection risk affect the expected returns and risk profiles of MEV extraction strategies.

The theoretical model involves solving a Bayesian Nash equilibrium with affiliated values, deriving a piecewise equilibrium bid function, and characterizing the builder's revenue-maximization problem. The empirical component requires processing 2.2 million transactions, estimating γ(τ) from right-tail bribe plateaus using quantile binning, and performing revenue decomposition. The game-theoretic derivations (ODE for equilibrium bidding, Leibniz rule for revenue derivative) are mathematically non-trivial but standard in auction theory. The empirical estimation is moderately complex, requiring careful handling of thin-market types (liquidations) and right-tail concentration effects.

The paper uses the publicly available libmev dataset (same as companion paper [1]) with 2.2 million MEV bundle transactions from September 2024 to August 2025. The theoretical model is fully specified with closed-form equilibrium characterizations. Log-normal parameters (μ̂=1.102, σ̂=2.524) are reported. However, no code repository is explicitly linked, and some estimation details (e.g., exact binning procedure for right-tail plateaus) could benefit from additional documentation.