Illuminating the Dark Forest: Unveiling the Mysteries of MEV

With the surge of activities on the Ethereum chain and the evolution of its infrastructure, MEV has been regarded as the most dangerous part of the dark forest, directly impacting users' on-chain financial activities and experiences. This article aims to analyze the centralization and trust issues brought by the current mechanism from the perspective of Ethereum 2.0's block generation mechanism and the proposer-builder separation (PBS), as well as the contradictions with Ethereum's decentralized values.

The exacerbation of MEV is a double-edged sword. The positive effects include reducing price discrepancies in DEX and assisting with liquidation; the negative impact mainly harms users through sandwich trading. Therefore, the solutions for MEV are more about mitigating the negative effects rather than eradicating them. Currently, there are three main types of measures: auction mechanism improvements, consensus layer enhancements, and application layer advancements. These solutions will affect the MEV landscape to varying degrees, but some proposals do not effectively address the sandwich attack issues faced by users. More privacy pool technologies need to be introduced to protect the optional privacy of user transactions, which is worth experimenting with.

MEV, as a byproduct of mechanism design, cannot be entirely avoided and will become even more complex in the future. The implementation of new transaction types such as Layer 2 architectures and EIP-4337 account abstraction will bring more technical challenges and opportunities related to MEV.

This article explores potential solutions to mitigate the negative impacts of MEV and provides a comprehensive understanding of the pros and cons of current MEV solutions, hoping to illuminate the "dark forest" for users and researchers.

Ethereum 2.0

After the Merge, Ethereum adopts a POS mechanism to ensure network security, with block production relying on proof of stake. After the merge, Ethereum is divided into an execution layer and a consensus layer. Each Epoch represents a POS cycle, consisting of 32 slots of 12 seconds each.

The validator committee is randomly selected in each Epoch, and the validator proposing the block is randomly chosen from the committee, responsible for packaging and sorting transactions and producing blocks, while other committee members supervise the voting. The committee is reselected after each Epoch. The block proposer implements the execution load ( transaction status change ) and block proposal.

PBS Architecture

Validators, as block proposers, often lack the motivation to execute Payload( sorting and execute transactions) due to the high computational requirements. To address this issue, the PBS concept was proposed, separating block proposing from construction. Proposers are only responsible for validating the blocks and do not participate in construction. This fosters an open market where block proposers can acquire blocks from builders. Builders compete with each other to construct blocks, offering the highest fees to proposers, forming a "block auction."

PBS Sealed First Auction Model Process: Users submit transactions to the public Mempool via RPC, and multiple Builders find the best transaction order to generate profit-maximizing blocks, interacting with the Proposer through MEV-Boost Relayer. Builders submit bids to the Relayer, and the Relayer submits multiple block headers and bids to the Proposer, who usually adopts the highest bid. The Relayer implements the MEVBoost specification, regulating the bidding interaction between Builders and Proposers. The entire process is information-closed, with the Relayer only submitting block headers to the Proposer to ensure censorship resistance.

Types of Participants and Game Theory under PBS

The main participants are Builder, Relayer, Proposer, MEVbot(Searcher).

Builder

The Builder is responsible for constructing block content, and the MEV-Boost technology gives it an advantage in bidding. Builders can directly review transactions, which has been criticized, especially after the OFAC released its guidelines, leading to a large number of Builders participating in OFAC Compliant activities. Although the proportion of reviewed blocks has recently declined, Builders still have a direct role in transaction reviews.

Currently, the share of the Builder market is gradually expanding with beaverbuild.org, which does not require review, everything is profit-oriented.

Searcher

Searchers collaborate with Builders to form Dark Pools or Private Pools, where Searcher transactions are only visible to specific Builders, allowing for maximized profit from MEV transactions. Theoretically, if a Builder engages in malicious behavior, it will lead Searchers to choose other Builders, reducing market share. Therefore, Builders will consider the implicit costs of malicious actions.

Searchers are divided into two main categories: CEX-DEX( off-chain) arbitrage and DEX, mezzanine, and liquidation( pure on-chain). Wintermute occupies the first position in CEX-DEX arbitrage. Pure on-chain MEV shows a trend towards studioization, with jaredfromsubway.eth holding a 37.2% market share, specializing in sandwich attacks. Many Searchers send order flow to the top three Builders to maintain ecological influence.

Relayer

Relayer集合竞价,向Proposer提交区块头和价格。Proposer签名后,Relayer释放全部交易内容。Relayer作为无经济激励的第三方获得极大信任,存在潜在风险。历史上出现过Ultrasound Relayer漏洞导致Proposer提取超2000万美元MEV的事件。

In the Relayer market share, the proportion of pure MAX Profit Builders has gradually increased since the Merge, making it difficult to control MEV artificially. Relayers face the issue of lacking economic incentives, and Blocknative has exited research and development in this direction. Currently, they rely on the MEVBoost specifications from Flashbots, while the Ethereum community is exploring the integration of PBS at the protocol level.

Proposer

The proposer randomly selects from validators, possessing execution load capacity, but tends to outsource. This can easily lead to vertical cooperation with builders, and the MEV-boost relayer hopes to act as an intermediary to reduce direct communication. Mining pools and LSD validator pools have strong scale effects, and the validator pool shows a trend toward centralization.

Lido holds a 28.7% market share, followed by Coinbase and Ether.fi. In the past, when MEV-BOOST PBS was not implemented, the Proposer had to take on the Builder tasks, but most gave up the ability to execute transaction ordering and instead outsourced the execution load.

User

Users are in the most vulnerable position, and transactions are placed in the Mempool, making it easy for MEV bots to profit. However, MEV also has positive effects, such as reducing DEX slippage and price differences. To avoid being monitored, there are RPC node providers that help users place transactions in non-public Mempools. OFA( order flow auction) is a novel way to compensate users for MEV profits by collaborating with Searchers.

Currently, the proportion of users using private order flow is about 10%, mainly limited by the cost of education and the complexity of operations.

Summary

Under the PBS architecture, after the introduction of MEV-BOOST, Builders and Searchers tend to cooperate, and various links show a trend of centralization, introducing multi-party trust issues that contradict Ethereum's vision of decentralization. The Ethereum community is currently discussing three proposals:

1. SUAVE Technology: Increase transaction transparency and lower the trust threshold of Searchers towards Builders.

2. Enshrined PBS: Incorporate PBS at the protocol level, eliminating reliance on Relayer.

3. Decentralized AVS: such as SSV, Lido has already partnered with it.

Current State of MEV

Currently, the main on-chain MEV includes arbitrage, sandwich attacks, and liquidations. Arbitrage profits are the highest, with MEV bots earning a profit of 2.6 million dollars in the past 30 days. The average profit from a single sandwich attack is 0.8 dollars, with a total profit of approximately 880,000 dollars in the past 30 days.

The positive effects of MEV include reducing DEX price differences and assisting with liquidation; the negative impact mainly involves user sandwich trading losses. In the short term, an increase in MEV activities will lead to higher Gas fees, affecting user experience.

During the transition of Layer 2 architecture, cross-chain MEV issues have also emerged. The account abstraction and new transaction types introduced by EIP-4337 will have a significant impact on the MEV landscape.

Potential MEV Complexity in Layer 2 Architecture Design

Future large-scale on-chain arbitrage activities will shift towards more complex multi-chain cross-chain MEV. Layer 2 mainly focuses on improvements for sorters, and cross-chain bridges are essential products. Searchers can help mitigate fragmented liquidity between Layer 2s, but the cross-chain bridge experience and security still need improvement.

Potential MEV of EIP-4337

EIP-4337 introduces account abstraction and a new transaction type called User Operation. Bundler is similar to Builder and can work with Searcher to reorder User Operations to capture MEV. The specifications for User Operations may vary across different chains, increasing the technical threshold for cross-chain MEV.

Exploration Direction for MEV Mitigation

Ethereum is considering using the e-PBS protocol-level improvement to avoid relying on third-party MEV-Boost relayers. Currently, MEV is difficult to fundamentally resolve, and more efforts are focused on alleviating negative impacts and enhancing positive effects. The main exploration directions are: protocol level, application level, and auction mechanism.

SUAVE

SUAVE aims to mitigate the negative effects of MEV, guiding MEV towards decentralization and transparency. By building a new blockchain SUAVE, it integrates the MEVM virtual machine to run EVM smart contracts. It lowers the barriers to creating MEV applications, enhances competition between mechanisms, and brings decentralization and transparency.

OFA

OFA( order flow auction ) improves the auction mechanism. The order initiator sends the order to OFA, and OFA selectively discloses information. Bidders place bids to obtain order flow information and provide payment prices. Private order flow is only visible to bidders, introducing market competition to make MEV more transparent and reduce user losses.

Private Trading Pool Encryption

Private trading pool encryption schemes include MPC, TSS, VDF, ZKP, etc., each with its own advantages and disadvantages. Notable projects include Shutter Network using TSS and Automate Network using MPC and ZKP.

Execution Tickets

Justin's proposed consensus layer improvement plan introduces the Tickets market. Obtaining a Ticket allows for the proposal of block execution in future periods. Blocks are divided into execution and proposal types, with execution blocks requiring a Ticket. This mechanism increases the deflationary pressure on ETH, but may give rise to multiple MEV issues, and the problem of user MEV attacks remains unresolved.

e-PBS

e-PBS( enshrined PBS) incorporates PBS into the consensus level, no longer relying on third-party solutions. The goal is to achieve trust-minimized PBS, capturing and distributing MEV through in-protocol mechanisms. The Relayer role is eliminated, with Builders bidding to write consensus layer code to Proposers.

Introducing PTC( Payload-Timeliness Committee ) supervises Builders to timely release payloads. Complete blocks are assembled from empty CL Blocks and Payloads. e-PBS solves the transparent auction solution without the need to trust third parties, separating the consensus layer and execution layer, improving efficiency. However, there are still some issues that need to be discussed.

PEPC

The AVS component PEPC( proposed by EigenLayer, which is protocol-enforced proposer commitments), aims to address the trust issues of Relayers. It requires Proposers to accompany CL block submissions with a PEPC signature, which the Builder verifies before executing the payload, thereby introducing a trust mechanism within the protocol.

![Illuminating the Dark Forest: Unveiling the Mystery of MEV](