Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MEV in Multiple Concurrent Proposer Blockchains

Steven Landers, Benjamin Marsh

TL;DR

This work formalizes maximal extractable value (MEV) in multiple concurrent proposer (MCP) blockchains, where data availability precedes final ordering and concurrency creates new MEV channels. It develops a hazard-normalized delay model with a delay envelope $M(\tau)$ and an immediate inclusion threshold $\tau^\dagger$, enabling analysis of censorship, duplication, timing, auctions, and inter-block MEV. The authors identify MCP-specific MEV channels such as same-tick duplicates and PoA-latency races, and propose two mitigation strategies: duplicate-aware tip-splitting and a deterministic Priority-DAG Merge Scheduler (PDM) that preserves dependencies while reducing latency races. They show that PoA latency governs stealability, and that with careful protocol choices (e.g., duplicate-aware payouts and PDM), MCP MEV can be controlled without centralized builders, with extensions to encryption and TFMs as avenues for further reduction.

Abstract

We analyze maximal extractable value in multiple concurrent proposer blockchains, where multiple blocks become data available before their final execution order is determined. This concurrency breaks the single builder assumption of sequential chains and introduces new MEV channels, including same tick duplicate steals, proposer to proposer auctions, and timing races driven by proof of availability latency. We develop a hazard normalized model of delay and inclusion, derive a closed form delay envelope \(M(τ)\), and characterize equilibria for censorship, duplication, and auction games. We show how deterministic priority DAG scheduling and duplicate aware payouts neutralize same tick MEV while preserving throughput, identifying simple protocol configurations to mitigate MCP specific extraction without centralized builders.

MEV in Multiple Concurrent Proposer Blockchains

TL;DR

This work formalizes maximal extractable value (MEV) in multiple concurrent proposer (MCP) blockchains, where data availability precedes final ordering and concurrency creates new MEV channels. It develops a hazard-normalized delay model with a delay envelope and an immediate inclusion threshold , enabling analysis of censorship, duplication, timing, auctions, and inter-block MEV. The authors identify MCP-specific MEV channels such as same-tick duplicates and PoA-latency races, and propose two mitigation strategies: duplicate-aware tip-splitting and a deterministic Priority-DAG Merge Scheduler (PDM) that preserves dependencies while reducing latency races. They show that PoA latency governs stealability, and that with careful protocol choices (e.g., duplicate-aware payouts and PDM), MCP MEV can be controlled without centralized builders, with extensions to encryption and TFMs as avenues for further reduction.

Abstract

We analyze maximal extractable value in multiple concurrent proposer blockchains, where multiple blocks become data available before their final execution order is determined. This concurrency breaks the single builder assumption of sequential chains and introduces new MEV channels, including same tick duplicate steals, proposer to proposer auctions, and timing races driven by proof of availability latency. We develop a hazard normalized model of delay and inclusion, derive a closed form delay envelope \(M(τ)\), and characterize equilibria for censorship, duplication, and auction games. We show how deterministic priority DAG scheduling and duplicate aware payouts neutralize same tick MEV while preserving throughput, identifying simple protocol configurations to mitigate MCP specific extraction without centralized builders.

Paper Structure

This paper contains 18 sections, 25 theorems, 100 equations.

Table of Contents

  1. Introduction
  2. MEV
  3. MEV in MCP
  4. A model of MEV
  5. Censorship
  6. Tx stealing
  7. Tx auctions
  8. Timing games
  9. Inter-block MEV
  10. Spam
  11. PoA latency and inclusion
  12. Multi-submission externalities
  13. Inter-block ordering rule
  14. Future work
  15. Conclusion
  16. ...and 3 more sections

Key Result

lemma 1

Let $U_{\mathrm{mev}}(\alpha,\tau) = \frac{A k}{k+\lambda}\!\left(1-e^{-(k+\lambda)\alpha}\right) + \tau\, e^{-\lambda\alpha}$ and define $M(\tau):=\sup_{\alpha\ge 0} U_{\mathrm{mev}}(\alpha,\tau)$. Then $M(\tau)$ is continuous and strictly increasing with $M(0)=\tfrac{A k}{k+\lambda}$ and $\lim_{\t where $\alpha^\star(\tau)$ equals $\tfrac{1}{k}\ln\!\tfrac{A k}{\lambda\tau}$ if $\tau< A k/\lambda

Theorems & Definitions (45)

  • lemma 1: Drop vs keep cutoff
  • lemma 2: Closed form of $M(\tau)$
  • lemma 3: Maximizer
  • lemma 4: Symmetric cutoff
  • lemma 5
  • proposition 1: Steal threshold
  • lemma 6: Symmetric mixed equilibrium of steal attempts
  • corollary 1: Anti-steal via equal tip-splitting
  • proposition 2: Myerson reserve under i.i.d. values
  • lemma 7: Auction vs internalization cutoff
  • ...and 35 more