Crypto Gloss

PBS (Proposer-Builder Separation)

Proposer-Builder Separation (PBS) is an architectural design for Proof-of-Stake block production that formally separates two roles that were previously unified in a single entity: the block proposer (the validator who is randomly selected by the Ethereum consensus protocol to create and attest to the next block) and the block builder (a specialized entity who constructs the most economically valuable block by optimally ordering transactions and capturing MEV), connecting them through a relay and a sealed-bid auction where the proposer simply selects the highest bid without seeing block contents, preventing MEV-driven centralization of validators while enabling a competitive marketplace for block construction efficiency. PBS emerged as the solution to a fundamental tension in post-Merge Ethereum: MEV extraction requires sophisticated software and capital to maximize, which would naturally push staking toward large, sophisticated validator operations — centralizing Ethereum consensus — unless the MEV work could be outsourced to specialists without requiring validators to trust them.

The Problem PBS Solves

MEV Centralization Pressure

Before PBS, the entity that builds a block also proposes it. In a world with significant MEV:

“`

Sophisticated validator firm:

  • Runs proprietary MEV bots
  • Captures $10,000 MEV per block on top of base reward
  • Returns: ~$20K/day on 10,000 ETH staked

Solo home validator (no MEV optimization):

  • Uses basic client, takes transactions by gas price
  • Captures near-zero MEV
  • Returns: ~$12K/day on 10,000 ETH staked

Result: MEV-aware validators earn 40%+ more →

Solo stakers can’t compete →

Staking centralizes toward sophisticated operators

“`

PBS breaks this link: even a home validator running a Raspberry Pi can earn the same MEV revenue as a professional firm, because they don’t build the block — they just choose the highest bid from competitive builders.

The Commit-Reveal Problem

PBS requires a “commit-reveal” scheme to prevent validators from stealing MEV from builders:

  1. Builder commits to a block header + bid (validator can see bid amount and block hash, but not contents)
  2. Validator signs the header if bid is highest
  3. Builder reveals full block (validator is now committed — they can’t change the block or front-run it)

Without this scheme, a validator who sees full block contents could steal the MEV themselves (by replacing the builder’s transactions with their own) and refuse to pay the builder.

Current Implementation: MEV-Boost (External PBS)

PBS is not yet natively implemented in the Ethereum protocol. The current version — called external PBS or ePBS precursor — is implemented via MEV-Boost, a middleware client developed by Flashbots:

“`

┌─────────────────────────────────────────────────────┐

│ MEV-Boost Flow │

│ │

│ Builders → submit blocks + bids → Relays │

│ Relays → verify bids, escrow blocks │

│ Validators → query relays via MEV-Boost │

│ → accept highest bid │

│ → sign block header │

│ → relay releases full block │

│ → validator proposes block │

└─────────────────────────────────────────────────────┘

“`

MEV-Boost adoption: Within weeks of the Merge, 80%+ of Ethereum validators adopted MEV-Boost. By 2024, ~90%+ of blocks are built via MEV-Boost builders rather than locally.

The Three Roles

Proposer (Validator)

  • Does NOT build the block (in ePBS)
  • Queries relays via MEV-Boost to find the highest bid
  • Signs the block header of the winning bid
  • Earns: base issuance reward + priority fees + builder bid
  • Risk: signing a block header and then the builder not releasing the block (mitigated by relay design)

Builder

  • Sources transactions from: public mempool, private order flow, searcher bundles
  • Submits block + bid to relays
  • Earns: MEV extracted minus relay fees minus bid paid to validator
  • Risk: validator not signing their block (rare; validators lose money if they don’t accept highest bid)

Relay

  • Receives builder blocks, verifies bids, holds blocks in escrow
  • Releases blocks only after validator signs header
  • Earns: fees from builders (Flashbots charges small basis points per ETH in bids)
  • Current relays: Flashbots, BloXroute, Agnostic, Ultra Sound

The relay trust assumption is the current weak point — validators must trust that relays won’t lie about bids or withhold blocks. This is why enshrined PBS is desirable.

Enshrined PBS (ePBS)

The Ethereum research community is working on enshrined PBS — native protocol support for PBS that eliminates the relay trust assumption:

  • How it works: The proposer-builder auction is implemented directly in the consensus layer, using cryptographic commitments instead of trusted relays
  • Status: In Ethereum research as of 2025; part of the longer-term roadmap
  • Benefits over MEV-Boost:
    No relay trust required
    Stronger censorship resistance (builders must include specific transactions)
    Better MEV redistribution mechanisms possible (e.g., PEPC — Protocol-Enforced Proposer Commitments)

Related proposals include FOCIL (Fork-Choice enforced Inclusion Lists) — a mechanism to force builders to include all transactions that validators have attested to, preventing censorship.

PBS and Censorship

PBS’s practical implementation (MEV-Boost) created an unexpected censorship vector:

  • Flashbots relay is operated by a US-based company
  • After OFAC sanctioned Tornado Cash in August 2022, Flashbots relay began excluding sanctioned transactions
  • Flashbots controls the majority relay market share → majority of blocks censored those transactions
  • Peak: ~73% of blocks OFAC-compliant (October 2022)

This demonstrated that external PBS via relays reintroduces a centralization point for censorship. Enshrined PBS with inclusion lists would prevent this by making censorship costly (validators who don’t include required transactions are penalized).

PBS vs. SUAVE

Flashbots has also proposed SUAVE (Single Unifying Auction for Value Expression) — a more radical redesign:

  • A dedicated blockchain for MEV auctions
  • Allows cross-domain MEV (across Ethereum + L2s + other chains)
  • Aims to democratize order flow away from dominant builders

SUAVE is an ambitious long-term vision; the near-term focus remains on enshrined PBS for Ethereum mainnet.

History

  • 2020: Vitalik Buterin and Ethereum researchers propose PBS as long-term design direction
  • 2021: “PBS” formalized in EIP discussions; Flashbots begins building PoW MEV-Geth
  • September 2022: The Merge + MEV-Boost launch simultaneously; external PBS deployed day-one
  • August–October 2022: OFAC sanctions Tornado Cash; Flashbots relay censors transactions; censorship controversy peaks
  • October 2022: Flashbots relay open-sourced; competing relays emerge
  • 2023: Enshrined PBS research accelerates; FOCIL and related proposals enter Ethereum research
  • 2024: MEV-Boost at ~90% adoption; builder concentration concerns grow; attester-proposer separation proposals circulate
  • 2025: ePBS targeting inclusion in post-Pectra Ethereum upgrades; relay censorship largely mitigated by validator diversity

See Also