Block builders are the market makers. Validators outsource block construction to specialized builders like Flashbots and bloXroute. This creates a single-point auction where builders compete to pay the validator, not a distributed market of users.
Ethereum Block Auctions Are Not Free Markets
Proposer-Builder Separation (PBS) creates a controlled, permissioned auction for block space, concentrating power among a few builders and relays. This is the antithesis of a free market and a critical flaw in Ethereum's MEV infrastructure.
The Centralized Auction Illusion
Ethereum's block space auction is a centralized price-setting mechanism that fails as a free market.
MEV extraction precedes user access. Builders fill blocks by maximizing extractable value (MEV) via arbitrage and liquidation bots. User transactions are residual filler, creating a two-tiered priority system that distorts pricing.
PBS enshrines, not solves, centralization. Proposer-Builder Separation (PBS) formalizes this builder role. It mitigates validator-level MEV but shifts centralization pressure to a builder cartel, as seen in Flashbots' dominant market share.
Evidence: Over 90% of Ethereum blocks are built by three entities. The 'auction' is a bilateral negotiation between a validator and a monolithic builder bundle, not a free-for-all.
Thesis: PBS Creates a Permissioned Cartel, Not a Free Market
Proposer-Builder Separation centralizes power into a small, permissioned group of builders, replacing a decentralized auction with a private negotiation.
PBS is a private auction. The free market for block space exists between users and builders, not builders and proposers. The builder-proposer interface is a permissioned relay network where a handful of entities like Flashbots and bloXroute control the flow of blocks.
Builders are the new validators. The economic power shifts from the 1M+ ETH staked to the ~5-10 elite builders who win the majority of blocks. This creates a cartel of capital where only those with massive MEV extraction capabilities and exclusive order flow can compete.
Relays are the gatekeepers. They enforce censorship resistance lists and attest to block validity, becoming trusted third parties in a system designed to be trust-minimized. This architecture mirrors the permissioned validator sets of chains like Solana or BSC, not Ethereum's permissionless ideal.
Evidence: Over 90% of Ethereum blocks are built by three entities. The builder market's Gini coefficient exceeds 0.95, indicating extreme centralization. This is the proposer-builder cartel in practice.
The Mechanics of Control: How the 'Market' is Rigged
Ethereum's Proposer-Builder Separation (PBS) created a new power dynamic where block builders, not validators, control transaction ordering and MEV extraction.
The Problem: The Builder Monopoly
A few dominant builders like Flashbots and bloXroute control over 80% of blocks. This centralizes the power to censor transactions and extract MEV, turning the 'auction' into a private sale to the highest bidder.
- Centralized Censorship Risk: Builders can exclude OFAC-sanctioned addresses.
- Opaque Order Flow: Users have no visibility into how their transactions are being reordered for profit.
The Solution: SUAVE
Flashbots' SUAVE is an attempt to decentralize the builder role. It's a specialized chain for decentralized block building and cross-chain MEV.
- Decentralized Memory Pool: Aims to break the private mempool monopoly.
- Preference Auction: Users can express preferences (e.g., frontrunning protection) that builders must compete to fulfill.
- Execution Market: Creates a competitive landscape for block building, not just proposing.
The Problem: MEV-Boost as a Crutch
The dominant PBS middleware, MEV-Boost, outsources block construction. While it increased validator profits, it entrenched builder centralization and created systemic risks.
- Relay Trust Assumption: Validators must trust relays not to steal their blocks or censor.
- Liveness Dependency: If major relays fail, block production halts.
- Profit Maximization Trap: Validators are economically incentivized to use the highest-paying, most centralized builders.
The Solution: Enshrined PBS (ePBS)
Ethereum's long-term roadmap includes enshrining PBS directly into the protocol consensus layer, removing the need for trusted relays.
- Protocol-Level Auction: Block building rights are auctioned on-chain, verifiable by all.
- Eliminates Relay Risk: No more trusted third parties between proposers and builders.
- Credible Neutrality: The protocol itself becomes the auctioneer, reducing censorship vectors.
The Problem: Private Order Flow (PFOF)
Searchers and users send transactions directly to builders via private channels (e.g., Flashbots Protect RPC) to avoid frontrunning, creating a two-tiered system.
- Public Mempool is Toxic: Remaining public transactions are heavily exploited.
- Data Asymmetry: Builders with exclusive order flow have an insurmountable advantage.
- Fragmented Liquidity: The market for block space is no longer unified or transparent.
The Solution: Fair Sequencing Services
Projects like Astria and Espresso are building decentralized sequencers that provide fair transaction ordering (e.g., first-come, first-served) before blocks are built.
- Time-Based Fairness: Reduces the value of latency races and frontrunning.
- Decentralized Censorship Resistance: No single entity controls the transaction queue.
- Rollup Integration: Provides a neutral base layer for L2s like Arbitrum and Optimism.
Builder & Relay Market Share: The Concentration Problem
A comparison of the dominant players in the Ethereum PBS ecosystem, highlighting the centralization of block production and validation power.
| Metric / Feature | Flashbots (MEV-Boost) | bloXroute | Titan Builder | Ultra Sound Relay |
|---|---|---|---|---|
Builder Market Share (30d Avg) | 38.2% | 17.5% | 14.8% | 9.1% |
Relay Market Share (30d Avg) | 42.5% | 18.3% | N/A | 12.7% |
Top 3 Builders Combined Share | 70.5% | 70.5% | 70.5% | 70.5% |
Censorship Resistance (OFAC Compliance) | ||||
Private Orderflow Integration | ||||
Avg. Block Value Extracted (ETH) | 0.08 ETH | 0.07 ETH | 0.06 ETH | 0.05 ETH |
Relay Latency (P99) | < 500ms | < 300ms | N/A | < 700ms |
Open Source Relay Client |
Why This Isn't Just a 'Temporary' Problem
Ethereum's block auction model is a structurally flawed market, not a temporary inefficiency.
The auction is broken. A true free market requires many independent buyers and sellers. In Ethereum's block space, the seller is a single monopolist (the block builder), and the buyers (searchers) are a small, collusive oligopoly.
MEV is the primary commodity. The auction's real product is MEV extraction rights, not simple transaction ordering. This creates perverse incentives where builders like Flashbots and bloXroute optimize for maximal extractable value, not user fairness.
Proposer-Builder Separation (PBS) formalizes this. PBS doesn't solve the problem; it institutionalizes the builder monopoly. The separation creates a new, unregulated financial layer where block building is a winner-take-all game dominated by specialized firms.
Evidence: Builder market share. Post-PBS, the top three builders (e.g., Titan Builder, beaverbuild) consistently produce over 80% of blocks. This is centralization by design, not a temporary phase of market immaturity.
Steelman: "But It's More Efficient!"
The 'efficiency' argument for Ethereum's block auction is a distraction that confuses speed for market fairness.
Efficiency is a red herring. The argument that a centralized auction is 'more efficient' than a free market is technically true but economically irrelevant. It confuses transaction ordering speed with price discovery. A single actor can order transactions faster, but this eliminates the competitive forces that determine the true cost of block space.
Centralization creates hidden costs. The proposer-builder separation (PBS) model, designed to mitigate this, has failed. Builders like Flashbots and bloXroute now dominate, creating an oligopoly that extracts maximal value. The efficiency gained in milliseconds is dwarfed by the economic rent extracted from users through MEV.
Compare to real markets. A stock exchange like the NYSE is not 'inefficient' because it has many participants; its liquidity pools and order books are the mechanism for fair price discovery. Ethereum's current auction is the equivalent of letting one high-frequency trader set all prices for a minute, which is efficient for them but predatory for everyone else.
Evidence: Builder market share. Data from mevboost.pics shows the top three builders consistently control over 80% of blocks. This concentration proves the auction is not a free market but a captured system where efficiency serves the cartel, not the network.
The Systemic Risks of a Non-Free Market
The current block-building process is a cartelized auction, not a free market, creating systemic risks for users and the chain's long-term health.
The Problem: The Builder Cartel
Block production is dominated by a few centralized builders like Flashbots, BloXroute, and Titan. This creates a single point of failure and censorship.\n- >90% of blocks are built by a handful of entities.\n- Censorship risk: Builders can exclude transactions from OFAC-sanctioned addresses.
The Problem: Extractive MEV
The lack of a free market allows searchers and builders to extract maximum value from users via arbitrage, sandwich attacks, and liquidations.\n- $1B+ annually in extracted MEV, primarily from users.\n- Sandwich attacks on DEX trades remain a persistent, costly threat.
The Solution: Credible Neutrality
A truly free market requires credibly neutral infrastructure. This means permissionless participation and commitment to inclusion.\n- PBS (Proposer-Builder Separation) must be fully realized.\n- SUAVE and other decentralized block-building initiatives aim to break the cartel.
The Solution: Intents & Private Order Flow
Shifting from transactions to intents (as seen in UniswapX and CowSwap) moves competition to the solver layer, protecting users.\n- User gets best price across all liquidity sources.\n- MEV is internalized as competition between solvers, not extracted from users.
The Problem: L2 Fragmentation
Each rollup (Arbitrum, Optimism, Base) runs its own mini-MEV auction, fragmenting liquidity and security. This creates cross-domain MEV arbitrage opportunities.\n- Inefficient capital locked in bridges.\n- Complex risks for protocols spanning multiple chains.
The Solution: Shared Sequencing
A decentralized, shared sequencer network (like Astria, Espresso) can provide atomic cross-rollup execution and fair ordering, restoring a unified free market.\n- Atomic composability across rollups.\n- MEV redistribution to rollups and users, not just builders.
The Path Forward: Enshrined PBS and SUAVE
Current block auctions are inefficient, opaque markets that enshrined Proposer-Builder Separation and SUAVE aim to fix.
Ethereum's block space auction is not a free market. It is a first-price sealed-bid auction where builders submit blind bids, leading to chronic overpayment and MEV leakage. This inefficiency is a direct subsidy to validators at the expense of users and sophisticated builders.
Enshrined Proposer-Builder Separation (PBS) is the protocol-level solution. It formally separates block building from block proposing, creating a competitive builder market. This forces builders to compete on execution quality, not just their ability to win a blind auction, directly reducing extractable value.
SUAVE (Single Unified Auction for Value Expression) is the complementary execution layer. It is a specialized mempool and decentralized block builder that aggregates user intents across chains. Unlike opaque private mempools, SUAVE's open auction for order flow aims to democratize MEV capture, similar to the intent-based model of UniswapX or CowSwap.
The combined system creates a real market. Enshrined PBS provides the trust-minimized framework, while SUAVE provides the competitive execution venue. This moves value from validators back to users and applications, fixing the core economic flaw in today's block production.
TL;DR for Protocol Architects
Ethereum's block space is a private, permissioned auction, not a free market, creating systemic inefficiencies and centralization vectors.
The Problem: Private Order Flow
Builders win blocks by bidding in a private, off-chain auction (PBS). This creates a two-tiered market where searchers and MEV bots pay builders directly, bypassing the public mempool and creating information asymmetry.\n- Centralization Risk: Top builders like Flashbots and Titan control ~80% of blocks.\n- Opaque Pricing: Users cannot see the true cost of inclusion, only the winning bid.
The Solution: Credible Commitments
Protocols enforce fair access through cryptographic commitments. Builders must commit to block content before learning if they won, preventing last-second manipulation.\n- Enshrined PBS (ePBS): Aims to formalize this at the protocol level.\n- SUAVE: Aims to decentralize block building itself.\n- Key Benefit: Reduces builder ability to censor or extract value via timing games.
The Problem: MEV is the Real Price
The "base fee" is a distraction. The true cost of block space is MEV (Maximal Extractable Value). Searchers arbitrage this value and bid it to builders, who then pay validators.\n- User Pays Twice: Once in gas, again in extracted MEV (e.g., sandwich attacks).\n- Inefficient Allocation: Value flows to intermediaries (Jito, bloXroute) instead of users or the protocol.
The Solution: Intents & Order Flow Auctions
Shift from transaction execution to outcome fulfillment. Users submit intents (desired state changes), and solvers compete to fulfill them optimally.\n- UniswapX, CowSwap: Aggregate user flow and auction it to solvers.\n- Key Benefit: Returns MEV value to users via price improvement and protects against frontrunning.\n- Architecture Shift: Requires new infrastructure like intent signaling layers and solver networks.
The Problem: Validator-Builder Collusion
Proposer-Builder Separation (PBS) is incomplete. Validators can still run their own builders (merged roles), creating a vertical monopoly. The builder with the closest relationship to the winning validator has an inherent advantage.\n- Trust Assumption: Relies on validators to honestly select the highest bid.\n- Cartel Formation: Large staking pools can internalize MEV capture, squeezing out independent builders.
The Solution: Enshrined PBS & Distributed Trust
Fully separate roles at the consensus layer and distribute trust. ePBS bakes the auction into the protocol, making bid selection verifiable and slashing misbehavior.\n- DVT (Distributed Validator Technology): Splits validator keys across nodes, reducing single-entity control.\n- Key Benefit: Creates a credibly neutral playing field where the best block, not the best relationship, wins.
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.