Summary
Proposer-Builder Separation (PBS) via MEV-Boost decouples block construction from block proposal, letting specialized builders compete for the right to fill each slot. Over 90% of Ethereum blocks now go through MEV-Boost, and the system captures ~400% more value than local builds — but has developed structural gaps in economics, robustness, performance, and services that have become increasingly urgent as of 2026.
Architecture
The Pipeline
Searchers → Builders → Relays → Proposers (validators)
- Searchers identify MEV opportunities and submit bundles to builders.
- Builders assemble full blocks that maximize total value (base fees + tips + MEV).
- Builders submit bids to one or more relays, which hold the full block body.
- Proposers (validators) see only block headers and bids via MEV-Boost; they select the highest bid without seeing the block contents until after they commit.
- The relay acts as trusted escrow: it verifies the block is valid and the payment is correct before releasing the header to the proposer and the full block to the network.
Why Relays Exist
The relay solves the simultaneous-reveal problem: builders won’t reveal their blocks (giving away MEV) unless payment is guaranteed; proposers won’t commit unless the block is valid. Relays are trusted intermediaries that handle this atomically.
Market Concentration (2026 data)
| Builder | Market share |
|---|---|
| Titan Builder | ~47.6% |
| BuilderNet | ~26.0% |
| Others | ~26.4% |
The top 2 builders construct ~74% of all blocks. This concentration is driven primarily by exclusive order flow: builders with privileged access to high-value transaction flow (OFA providers, wallets, CEX-DEX arbitrage) win disproportionately.
Relay landscape: Flashbots, bloXroute, Ultra Sound, Titan, Aestus, and others. No relay has a dominant position, but all are US/Europe-based — validators in Asia face additional latency penalties when receiving high-value blocks.
Value Captured
- MEV-Boost blocks yield ~400% more ETH than locally-built blocks for proposers.
- Exclusive order flow can account for up to 84% of winning block fees in some blocks.
- This creates a flywheel: more flow → higher bids → more validators → more flow.
Four Structural Gaps (Blockspace Forum, Dec 2025)
The “Observation on Ethereum’s Blockspace Market” paper identified four systemic problems:
1. Economics
- Relays have no sustainable business model: they bear infrastructure costs but cannot charge builders without losing volume, and cannot charge proposers without losing validators.
- Builder market concentration is self-reinforcing via exclusive orderflow flywheel.
- Bid adjustment: ~5% of slots involve relay bid adjustment (e.g., the relay changes a builder’s bid); ~2% of total bid value is affected. This is a safety/trust issue with no protocol remedy today.
2. Robustness
- Three robustness properties Ethereum needs: censorship resistance, MEV distribution, and builder competition.
- A 400% PBS premium creates strong validator dependence on a healthy builder market — a fragile setup if builders collude or exit.
- No fallback if relays go down: validators miss the slot or fall back to local building (much lower value).
3. Performance
- Latency is a direct competitive advantage: builders closer to block producers win more often.
- Geographic concentration of relays (US/EU) disadvantages Asian validators.
- Timing games: builders delay bid submission to the last millisecond to capture fresher CEX prices, pushing MEV upward but introducing missed-slot risk.
4. Services
- Pre-confirmations require builder integration today — slow to roll out.
- No reliable mechanism for proposers to express commitments (FOCIL, pre-conf protocols need relay/builder integration).
- Builders can’t coordinate even when cooperation would benefit all.
Path to ePBS
MEV-Boost is extraconsensus: it relies on social trust and relay honesty. ePBS (EIP-7732) enshrines this split at the protocol level:
- Proposers commit to a builder’s payload hash at consensus time.
- The payload is released in the same slot (Execution Tickets variant) or a later reveal.
- Relays become optional or disappear; their escrow function is handled by the protocol.
- Scheduled for Glamsterdam (H1 2026).
Related Pages
- Relay Block Merging — Relay-level block enhancement post-auction
- Exclusive Order Flow and the Builder Flywheel — The order flow flywheel driving builder concentration
- — Builder delay strategies
- ePBS: Enshrined Proposer-Builder Separation (EIP-7732) — Protocol-level PBS
- BuilderNet and Decentralized Block Building — TEE-based shared building (Phase 1 toward SUAVE)
- MEV Supply Chain: Searchers, Builders, Relays, and Validators — Actor-level breakdown
Key Sources
- Three Years of PBS (Kubi M., Alex T., Kevin L., Justin D., Dec 2025) — structural gap taxonomy
- Economics Align the Incentives (2026) — exclusive orderflow flywheel analysis
- Ethereum: A Counterparty Without Counterparty Risk (2026) — robustness properties; 400% premium data
- Decentralized Building: wat do? (Flashbots, Feb 2026) — Phase 0–3 roadmap; SUAVE vision
- Blockspace Forum Workshop Cannes (Apr 2026) — relay merging, RLNC, sub-slots