MEV and Proposer-Builder Separation: How Ethereum Validators Actually Make Money

Running an Ethereum validator costs 32 ETH upfront, consumes electricity around the clock, and demands near-perfect uptime. What does a validator get in return? The answer is more complicated than "staking rewards." Since the Merge in September 2022, validator income has split into multiple streams: consensus layer attestation rewards, execution layer priority tips, and payments from specialized block builders who bid for the right to construct each block. That final stream is where maximal extractable value (MEV) enters the picture.

MEV has reshaped Ethereum's validator economics, created an entirely new class of infrastructure actors, and raised hard questions about censorship resistance. Understanding how it works is essential for anyone evaluating proof-of-stake security, Layer 2 design, or the long-term trajectory of decentralized networks.

What Is MEV and Why Does It Exist

MEV refers to the profit that can be extracted by reordering, inserting, or excluding transactions within a block. The concept was originally called "miner extractable value" under proof-of-work. After the Merge, the community adopted "maximal extractable value" since validators, not miners, now propose blocks.

MEV exists because transaction ordering matters. When a large swap moves the price on a decentralized exchange, someone who can place trades immediately before and after that swap captures a risk-free profit. When a lending position falls below its collateral threshold, the first transaction to liquidate it collects a reward. These opportunities are not bugs: they are structural features of any system where transaction ordering is discretionary.

Cumulative MEV extraction on Ethereum has exceeded $1.3 billion since the Merge, according to tracker estimates. Daily extraction ranges from $3 million to $20 million on quiet days, spiking to $40 million or more during periods of high volatility.

The MEV Supply Chain: From Searcher to Validator

The modern MEV pipeline involves four distinct actors, each performing a specialized function. This division of labor emerged through proposer-builder separation (PBS): the principle that the entity proposing a block should not be the same entity constructing it.

Searchers

Searchers are bots and algorithms that continuously monitor the public mempool and on-chain state for profitable opportunities. Common strategies include DEX arbitrage (exploiting price differences between trading venues), frontrunning and backrunning large trades, and liquidation of undercollateralized positions in lending protocols.

Searchers package their desired transactions into "bundles" and submit them to builders. Competition among searchers is intense: in competitive arbitrage scenarios, searchers may pay 90% or more of the captured value to builders as tips, retaining only a thin margin.

Builders

Builders receive bundles from multiple searchers, combine them with ordinary user transactions from the public mempool, and assemble complete, valid blocks optimized for maximum total revenue. Building blocks profitably requires sophisticated algorithms, low-latency infrastructure, and, increasingly, access to private transaction flows.

The builder role has become one of the most capital-intensive and technically demanding positions in the Ethereum ecosystem. A builder that can access exclusive order flows (private transaction streams from wallets, DEX frontends, or intent protocols) gains a significant advantage: those transactions are invisible to competing builders, enabling higher-value blocks.

Relays

Relays sit between builders and validators. When a builder constructs a block, it submits the block to one or more relays. The relay verifies the block is valid, holds the full transaction list in escrow, and passes only the block header and bid amount to the proposing validator. This prevents the validator from inspecting the block's contents, stealing the builder's MEV strategy, and constructing a competing block.

Major relays include Ultra Sound Money (approximately 34% of relay deliveries), Titan (24%), BloXroute (15%), and Aestus (10%). Relays operate as trusted intermediaries: if a relay fails to deliver the payload after the validator commits to the header, the validator misses its slot and earns nothing.

Proposers (Validators)

The proposing validator runs MEV-Boost, open-source middleware developed by Flashbots. During their assigned slot, the validator receives bids from registered relays and selects the highest bid. The validator does not see the block's contents: they simply sign the header and earn the bid amount deposited to their fee recipient address.

How MEV-Boost works in practice: Approximately 90% of Ethereum validators run MEV-Boost. Blocks built through MEV-Boost average 0.12 ETH per block compared to 0.04 ETH for locally built blocks: roughly a 3x revenue premium. This economic incentive makes opting out of MEV-Boost irrational for most validators.

How Validator Revenue Breaks Down

Ethereum validators earn income from three sources, each with different characteristics in terms of predictability, magnitude, and variance.

The base staking APR for Ethereum validators sits at approximately 2.8% as of mid-2026, with roughly 38.9 million ETH staked across approximately 897,000 active validators. MEV-Boost adds an estimated 0.3% to 0.8% additional APR, bringing the realistic all-in yield to 3.3% to 3.8% for well-operated nodes.

The ratio between consensus and execution layer rewards fluctuates significantly. During periods of intense DeFi activity (token launches, market crashes, liquidation cascades), execution layer rewards can temporarily exceed consensus rewards. In quiet markets, consensus rewards dominate at 90% or more of total income.

The Lottery Problem

A solo validator with 32 ETH proposes a block roughly once every 4 to 5 months on average. MEV payouts are concentrated in block proposals, not attestations. This means MEV income is extremely lumpy for individual validators: most slots yield nothing, while an occasional slot during a volatility spike could pay several ETH.

This variance pushes validators toward staking pools and liquid staking providers, which smooth MEV income across many validators. Lido, Coinbase, and similar services aggregate thousands of validators, socializing the MEV windfall across all stakers. This dynamic further concentrates effective control over Ethereum's validator set.

Builder Market Concentration: An Oligopoly Problem

The most pressing concern in Ethereum's MEV landscape is builder centralization. Academic research covering September 2023 through August 2025 identified four distinct eras of builder market evolution, culminating in what researchers term an "oligopoly" phase where three builders consistently controlled roughly 90% of block production.

Titan Builder alone booked over $112 million in block-building revenue in the first half of 2026. In one extreme case in July 2026, Titan extracted $1.8 million from a single swap where a trader swapping 1,126 ETH lost 99.3% to same-block arbitrage.

Why Exclusive Order Flow Drives Concentration

The primary engine of builder concentration is exclusive order flow (EOF): private transaction streams that only one builder can access. Research has identified 75 distinct EOF arrangements that account for over 70% of total trading-related builder revenue.

When a wallet provider, DEX frontend, or intent protocol routes transactions exclusively to one builder, that builder gains information and transaction inventory that competitors cannot match. The result is a self-reinforcing cycle: builders with more EOF produce more profitable blocks, win more auctions, attract more EOF agreements, and extend their market share.

The Beaverbuild Transition

In a significant development for builder decentralization, Beaverbuild (the historically dominant builder since 2022) retired its centralized block builder on May 6, 2025, and migrated entirely to BuilderNet: a decentralized block-building network jointly operated by Flashbots, Beaverbuild, and Nethermind. BuilderNet runs inside trusted execution environments (TEEs) and is designed to neutralize exclusive order flow advantages by processing transactions in an environment where no single operator can extract private information.

Censorship Risks and OFAC Compliance

Builder concentration raises direct censorship resistance concerns. If a small number of builders produce most blocks, those builders can selectively exclude transactions: whether due to regulatory pressure, commercial incentives, or government orders.

This concern became tangible in August 2022 when the U.S. Treasury's Office of Foreign Assets Control (OFAC) sanctioned Tornado Cash. In response, multiple relays began filtering transactions that interacted with sanctioned addresses. By October 2022, the percentage of blocks complying with OFAC sanctions lists peaked at approximately 95%: meaning nearly every block excluded transactions involving sanctioned contracts.

Recovery and Current State

The situation has improved significantly since that peak. As of September 2026, approximately 25% of MEV-Boost relay payloads route through OFAC-censoring relays, according to MEV Watch. The recovery was driven primarily by the growth of neutral relays (Ultra Sound Money, Agnostic, Titan, Aestus) that do not filter transactions based on sanctions lists.

The censorship arithmetic: Even if 25% of blocks censor a transaction, that transaction will still be included within a few minutes on average: it simply needs to wait for one of the 75% of non-censoring blocks. True censorship requires near-total builder or relay compliance, which becomes harder as the relay and builder landscape diversifies. The concern is less about permanent exclusion and more about creating a precedent where protocol-layer actors routinely make policy decisions about which transactions are valid.

The Road to Enshrined PBS: EIP-7732

The current MEV-Boost system works, but it operates entirely outside Ethereum's protocol. Relays are trusted third parties with no on-chain accountability. If a relay fails to deliver a payload, the validator loses its slot reward with no recourse. If a relay colludes with a builder, there is no protocol-level penalty.

EIP-7732 (enshrined proposer-builder separation, or ePBS) moves the builder auction into the Ethereum protocol itself. It is scheduled for inclusion in the Glamsterdam hard fork, targeting Q3 2026. As of May 2026, a stable multi-client devnet was running the latest ePBS specification.

How ePBS Changes the Game

EIP-7732 introduces several structural changes to how blocks are built and proposed:

• Builders become protocol-recognized entities, not just off-chain actors submitting bids through middleware
• A new Payload Timeliness Committee (PTC) of validators attests whether builders reveal their payloads on time
• Full execution payload validation is deferred to the next slot, separating consensus validation from execution validation
• Two complementary markets emerge: a trustless peer-to-peer market (gossip-based bidding) and an RPC market (private, similar to current MEV-Boost arrangements)

The most significant implication: relays become economically unnecessary. The protocol itself enforces builder commitments, eliminating the need for a trusted intermediary to hold blocks in escrow. Validators gain genuine optionality between local block building, trustless P2P bidding, and direct RPC arrangements with builders.

What ePBS Does Not Solve

Enshrining PBS at the protocol level addresses relay trust assumptions but does not directly solve builder concentration. Builders with superior algorithms, more exclusive order flow, and better latency will continue to win auctions under ePBS. The competitive dynamics shift (relay trust disappears), but the fundamental advantages driving concentration remain.

Additional proposals are needed to address builder centralization more directly. These include inclusion lists (forcing builders to include specific transactions), encrypted mempools (hiding transaction contents until after ordering), and block-level data availability improvements.

MEV in Layer 2 Context

MEV is not unique to Ethereum L1. Any system with discretionary transaction ordering creates MEV opportunities. Ethereum's rollups face their own MEV challenges: sequencers on optimistic and ZK rollups control transaction ordering within their domains, creating extraction opportunities that mirror L1 dynamics. Flashbots has developed Rollup-Boost, a sidecar for extending MEV-Boost capabilities to L2 sequencers.

The MEV problem scales with DeFi complexity. Chains with deep AMM liquidity, lending protocols, and complex smart contract interactions generate more MEV. This has implications for how alternative Layer 2 architectures are designed: systems that minimize or eliminate discretionary transaction ordering can avoid creating MEV extraction opportunities entirely.

Bitcoin Layer 2 designs face a fundamentally different MEV landscape. The MEV extraction risks on Bitcoin L2s are structurally lower because Bitcoin's scripting model does not support the complex DeFi interactions that generate most Ethereum MEV. Spark, for example, uses a statechain architecture where transfers involve key rotations rather than transaction ordering within blocks. There is no mempool to inspect, no transaction sequence to manipulate, and no AMM state to sandwich. The MEV surface area is effectively zero by design.

Comparing MEV Across Architectures

The degree of MEV exposure varies dramatically depending on the architecture of a given network. The following comparison highlights how different design choices affect MEV risk.

What This Means for Validators and Stakers

For individual stakers and institutions evaluating validator economics, several practical takeaways emerge:

• Running MEV-Boost is effectively mandatory for competitive returns: the 3x revenue premium makes opting out a measurable cost
• MEV income is highly variable and favors large staking operations that can smooth variance across many validators
• The shift to ePBS in Glamsterdam will change the operational requirements but not the fundamental economics: builders will still bid for block construction rights
• Builder concentration creates systemic risks that individual validators cannot mitigate on their own: this is a protocol-level challenge
• Validator yield will likely continue compressing as more ETH is staked, making MEV an increasingly important component of total returns

For a deeper look at how staking yields compare across proof-of-stake networks, see our analysis of restaking economics and risks. Developers interested in building on networks with different MEV profiles can explore the Spark SDK documentation for a Layer 2 architecture designed without MEV extraction vectors.

Looking Ahead

The MEV ecosystem is not static. Several developments will shape its trajectory through 2026 and beyond:

ePBS (EIP-7732) will remove relay trust assumptions when Glamsterdam ships, but builder concentration will persist until additional mechanisms like inclusion lists or encrypted mempools are deployed. BuilderNet represents an experiment in decentralized block building, but its 13% market share suggests the approach has not yet disrupted the dominance of vertically integrated builders like Titan.

MEV will remain a defining feature of Ethereum's economic model. It funds validator operations, incentivizes sophisticated infrastructure, and creates extraction risks for ordinary users. Whether the ecosystem can distribute MEV benefits more broadly while reducing its concentration effects will determine whether Ethereum maintains the decentralization properties that justify its architecture.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.