Free 30-min Web3 Consultation
Book Consultation
Smart Contract Security Audits
View Audit Services
Custom DeFi Protocol Development
Explore DeFi
Full-Stack Web3 dApp Development
View App Services
Free 30-min Web3 Consultation
Book Consultation
Smart Contract Security Audits
View Audit Services
Custom DeFi Protocol Development
Explore DeFi
Full-Stack Web3 dApp Development
View App Services
Free 30-min Web3 Consultation
Book Consultation
Smart Contract Security Audits
View Audit Services
Custom DeFi Protocol Development
Explore DeFi
Full-Stack Web3 dApp Development
View App Services
Free 30-min Web3 Consultation
Book Consultation
Smart Contract Security Audits
View Audit Services
Custom DeFi Protocol Development
Explore DeFi
Full-Stack Web3 dApp Development
View App Services
layer-2-wars-arbitrum-optimism-base-and-beyond
Blog

The Future of Proposer-Builder Separation (PBS) on Rollups

Ethereum's PBS model breaks on rollups. This analysis explains why Sequencer-Builder-Separation (SBS) is the necessary evolution, the critical challenges of latency and trustless auctions, and what it means for Arbitrum, Optimism, and Base.

introduction
THE ARCHITECTURAL IMPERATIVE

Introduction

Proposer-Builder Separation (PBS) is the inevitable architectural model for scaling rollup decentralization and censorship resistance.

PBS is inevitable for rollups. The current centralized sequencer model creates a single point of failure and censorship. PBS, pioneered by Ethereum's mev-boost, separates block proposal from block construction, enabling specialized builders like Flashbots SUAVE to compete.

The goal is credible neutrality. PBS prevents a single entity from controlling transaction ordering and MEV extraction. This is the foundational requirement for permissionless, trust-minimized rollups that must inherit Ethereum's security properties.

The challenge is implementation. Native in-protocol PBS, like Ethereum's ePBS, is complex. Rollups are adopting hybrid models, with projects like Arbitrum BOLD and Optimism's Fault Proof System exploring decentralized sequencing layers that incorporate PBS principles.

Evidence: Ethereum's post-merge adoption of mev-boost reached >90% validator participation, proving the model's viability. Rollups like Taiko are launching with PBS-inspired designs from day one.

thesis-statement
THE ARCHITECTURAL DIVIDE

The Core Argument: PBS ≠ SBS

Proposer-Builder Separation (PBS) on L1 is a market design, but on rollups it becomes a fundamental architectural split.

Ethereum PBS is optional. It's a market-based solution for MEV extraction, enforced via relays like Flashbots and builders like Titan. Rollup PBS is architecturally mandatory; the sequencer is the builder, and the L1 is the proposer.

SBS is a different problem. Shared Sequencing (SBS) solves atomic composability across rollups, a challenge for AltLayer and Espresso. PBS solves MEV centralization and censorship resistance within a single rollup's execution.

Rollups cannot outsource trust. An L1 proposer can trust a malicious builder because the L1 enforces correctness. A rollup's proposer (L1) must re-execute the builder's (sequencer's) block, making the builder's role purely for ordering.

Evidence: Arbitrum BOLD demonstrates this. Its dispute protocol forces the L1 to be the ultimate proposer, making any PBS implementation a permissioned whitelist, not a permissionless auction.

PROTOCOL DESIGN

PBS vs. SBS: The Hard Constraints

A comparison of Proposer-Builder Separation (PBS) and Sequencer-Builder Separation (SBS) architectural models for rollups, focusing on concrete trade-offs for infrastructure and economic security.

Constraint / MetricEnshrined PBS (e.g., Ethereum)Permissioned SBS (e.g., Espresso, Astria)Centralized Sequencer (Status Quo)

Builder Collusion Resistance

Censorship Resistance (L1-level)

MEV Extraction Surface

Builder-level auction

Sequencer-level auction

Sequencer internal

Time to Finality Impact

Adds 1 block (12s)

Adds < 1 block

Native rollup latency

Implementation Complexity

Protocol-level fork

Middleware integration

Native client feature

Required Trust Assumption

L1 consensus

Sequencer committee honesty

Single sequencer operator

Proposer/Sequencer Revenue

Bid from builder

Bid from builder

All fees + internal MEV

Adoption Timeline

Post-Danksharding (2025+)

2024-2025

Live today

deep-dive
THE ROLLUP FRONTIER

The Two Unsolved Problems of SBS

Proposer-Builder Separation on rollups introduces new, unresolved challenges around cross-domain MEV and trust-minimized settlement.

Cross-Domain MEV Extraction is the primary unsolved problem. Rollup sequencers are natural builders, but they operate in isolated domains. A sophisticated actor running a sequencer on Arbitrum and a builder on Ethereum can capture cross-domain arbitrage that a single-domain builder misses. This creates a new, more complex MEV supply chain that protocols like SUAVE aim to coordinate.

Trust-Minimized Settlement remains a critical gap. Ethereum's PBS relies on the protocol-enforced crediting of the proposer. Rollups lack this native mechanism, forcing reliance on off-chain agreements or centralized sequencer lists. Without a protocol-level slashing condition for builder payments, systems like Espresso or Astria must build complex reputation and bonding layers to prevent theft.

Evidence: The current model is fragmented. Optimism's RPGF funds research into decentralized sequencing, while Arbitrum's BOLD consensus focuses on fraud proofs, not PBS economics. No major L2 has implemented a production, trust-minimized SBS framework, proving the problem's complexity.

protocol-spotlight
THE FUTURE OF PBS ON ROLLUPS

Ecosystem Approaches: Who's Solving What?

Ethereum's PBS is a core scaling primitive, but its implementation on rollups is fragmented and incomplete. Here are the key battlegrounds.

01

The Problem: Centralized Sequencing is a Black Box

Most rollups run a single, centralized sequencer. This negates PBS's core value of separating block production from block proposal, creating a single point of failure and censorship.

  • No MEV Redistribution: All value is captured by the sequencer operator.
  • Opaque Operations: Users cannot audit transaction ordering or inclusion fairness.
  • Protocol Risk: Relies on the sequencer's altruism for liveness and correct execution.
>95%
Centralized
0%
MEV Rebate
02

The Solution: Permissionless Proposer Auctions (Espresso, Astria)

Decentralize the proposer role via a fast, leaderless auction. Rollups outsource block production to a competitive market of builders, while retaining execution sovereignty.

  • Shared Sequencing Layer: Projects like Espresso Systems and Astria provide a neutral marketplace for proposers.
  • Rollup Sovereignty: The rollup's DA and settlement layer remain the ultimate arbiters of chain state.
  • MEV Capture & Redistribution: Auction revenue can be directed to the rollup's treasury or its users, akin to EigenLayer's vision for Ethereum.
~1s
Auction Time
N Builders
Competitive
03

The Solution: Encrypted Mempools & MEV Mitigation (Shutter, Fairblock)

Prevent harmful MEV at the source by encrypting transactions until they are included in a block. This shifts the PBS dynamic from redistribution to prevention.

  • Threshold Encryption: Networks like Shutter Network use a distributed key shard system to blind transaction content.
  • Builder Obliviousness: Builders bid on encrypted bundles, unable to front-run or sandwich users.
  • Composability: Can be integrated with existing PBS designs like those from Flashbots (SUAVE) or Builder APIs.
~99%
MEV Reduction
+200ms
Latency Add
04

The Problem: L2 PBS Fragments Liquidity

Each rollup implementing its own PBS auction creates isolated MEV markets. This reduces builder capital efficiency and increases latency for cross-domain arbitrage.

  • Capital Silos: Builders must post bonds and manage liquidity across dozens of chains.
  • Inefficient Arbitrage: Opportunities between, say, Arbitrum and Optimism are harder to capture atomically.
  • Weaker Security: Smaller, isolated auctions are more susceptible to manipulation.
N Auctions
Fragmented
-70%
Efficiency
05

The Solution: Cross-Rollup PBS Hubs (SUAVE, Anoma)

Create a unified, domain-agnostic block building market. A single decentralized builder network can service multiple rollups, aggregating liquidity and intent.

  • Unified Preference Environment: Flashbots' SUAVE aims to be a decentralized block builder and encrypted mempool for all chains.
  • Intents-Based Flow: Users express desired outcomes (e.g., "swap X for Y at best rate across L2s"), which builders fulfill optimally, similar to UniswapX.
  • Maximal Extractable Value (MEV) becomes Maximal Extractable Coordination, capturing cross-domain efficiency gains.
1 Network
All Rollups
$B+
Liquidity Pool
06

The Solution: Enshrined PBS at the Settlement Layer (EigenLayer, EigenDA)

Bake PBS directly into the rollup stack's security model. Use Ethereum's cryptoeconomic security to enforce proposer decentralization and slashing conditions.

  • Restaking Security: Projects like EigenLayer allow ETH stakers to extend security to auxiliary services, including a decentralized proposer network for rollups.
  • Credible Neutrality: The proposer set is secured by Ethereum stake, not a rollup's native token, aligning incentives with the base layer.
  • Data Availability: Coupled with EigenDA or Celestia, this creates a fully decentralized, modular rollup stack.
$10B+
Security Pool
L1 Slashing
Enforcement
counter-argument
THE SIMPLE SOLUTION

Steelman: "Just Use a Token"

The most direct path to PBS on a rollup is a native auction token, but this introduces new economic and security trade-offs.

A native auction token is the simplest PBS mechanism. The rollup's sequencer role is permissionless, and block-building rights are auctioned in the native token, similar to Ethereum's MEV-Boost. This creates a clear economic alignment where builders must acquire and stake the token to participate, directly linking their success to the rollup's value.

This model centralizes economic risk. A token-based auction creates a single point of failure in the token's security and liquidity. If the token price crashes, the cost of attacking the sequencing auction plummets. This is a weaker security model than Ethereum's PBS, which uses ETH—a deeply liquid asset with established credibly neutral properties.

The builder market fragments. Unlike Ethereum's unified builder ecosystem, each rollup with its own token fractures builder capital and expertise. A builder must hold and manage risk across dozens of illiquid tokens instead of concentrating capital in ETH. This reduces competition and efficiency, potentially leading to higher user fees.

Evidence: The failure of early staking tokens like Polygon's MATIC for Heimdall validators demonstrates the operational overhead. Projects like Aevo and dYdX v4 use native tokens for sequencer selection, but their long-term security under adversarial market conditions remains untested compared to the $400B+ economic security of Ethereum's validator set.

risk-analysis
THE PBS FRAGILITY FRONTIER

The Bear Case: What Could Go Wrong?

Proposer-Builder Separation is the bedrock of modern block production, but its adaptation to rollups introduces novel systemic risks.

01

The Centralizing Force of Enshrined PBS

Rollups like Arbitrum and Optimism are exploring enshrined PBS, baking it into the protocol. This risks creating a permissioned, state-sanctioned builder market from day one, stifling the permissionless innovation seen on Ethereum.\n- Protocol Capture: Core devs become gatekeepers of builder entry.\n- Stagnant Design: Hard to fork or upgrade the auction mechanism post-launch.

0
Permissionless Builders
~100%
Protocol Control
02

MEV Cartels Go Cross-Chain

Dominant builders like Flashbots and Titan will extend their L1 strategies to rollups, creating vertically integrated MEV supply chains. This leads to value leakage from L2 users back to L1 searchers.\n- Extractive Bridges: Cross-domain MEV bundles become the norm.\n- L2 as a Colony: Rollup revenue gets siphoned by L1 entities, undermining sustainable fee markets.

$1B+
Annual Value Leakage
<10
Cartel Entities
03

The Liquidity Fragmentation Death Spiral

Sovereign rollups and high-throughput L2s will each host isolated builder markets. This fragments liquidity for block space auctions, making them inefficient and vulnerable to manipulation.\n- Thin Markets: Small, volatile auctions are easy to game.\n- Builder Exit: No economies of scale, leading to market failure and re-centralization.

100+
Fragmented Markets
-90%
Auction Efficiency
04

Data Unavailability as a Censorship Vector

Builders require timely data to construct profitable blocks. If sequencers or data availability layers like Celestia or EigenDA are unreliable, only a few well-connected builders can operate, creating a censorship-prone oligopoly.\n- Information Asymmetry: Privileged access to mempools and state data.\n- De-facto Censorship: Independent builders are priced out by latency.

~500ms
Advantage Window
3-5
Viable Builders
05

Regulatory Capture of the Builder Role

As PBS formalizes the 'block producer' role, it creates a clear regulatory target. Entities like Jump Crypto or Coinbase operating as builders could be forced to comply with KYC/AML on transactions, breaking censorship resistance.\n- OFAC-Compliant Blocks: The norm, not the exception.\n- Protocol-Level Sanctions: Compliance becomes a protocol requirement.

>50%
Blocks Censored
100%
KYC'd Builders
06

Complexity Collapse and Verification Paralysis

PBS adds a layer of cryptographic complexity and multiple rounds of communication. In a crisis, this slows down fault proofs and increases the risk of liveness failures. Provers for zkRollups may not keep pace with optimized builder outputs.\n- Verification Gap: Builders outrun provers, creating unsafe chain states.\n- Cascade Failure: A bug in the PBS mechanism halts the entire rollup.

10x
Attack Surface
Minutes
Recovery Time
future-outlook
THE ARCHITECTURE

The Path Forward: Hybrid Models and Shared Sequencing

The future of PBS on rollups involves hybrid models that blend centralized performance with decentralized guarantees, moving towards shared sequencing layers.

Pure decentralization fails at scale. A single, fully decentralized sequencer for a high-throughput rollup creates latency and inefficiency that users reject. The market demands centralized sequencing for performance, as proven by Arbitrum and Optimism's dominant market share.

Hybrid PBS models are inevitable. Rollups will adopt a builder market for block construction while retaining a decentralized, staked validator set for proposing and attesting. This separates the capital-intensive role from the latency-sensitive one, mirroring Ethereum's evolution.

Shared sequencing is the endgame. Protocols like Espresso Systems and Astria are building shared sequencing layers that multiple rollups use. This creates atomic cross-rollup composability and turns sequencing into a commodity, reducing individual rollup development overhead.

Evidence: Arbitrum and Optimism, which use centralized sequencers, command over 80% of rollup TVL. Their success validates the performance-first imperative that all future designs must address.

takeaways
THE FUTURE OF PBS ON ROLLUPS

TL;DR: Key Takeaways

Proposer-Builder Separation is migrating from Ethereum L1 to rollups, creating new MEV markets and existential design choices for rollup architects.

01

The Problem: In-House Builders Create Centralization

Most rollups today run a centralized, in-house sequencer that acts as both proposer and builder. This is a single point of failure and censorship, and it prevents an open MEV market from forming.\n- Centralized Control: The rollup team controls all transaction ordering and MEV extraction.\n- Stifled Innovation: No competition for block building efficiency or advanced PBS concepts like MEV smoothing.

1
Builder
100%
Control
02

The Solution: Permissionless PBS Auctions (à la Ethereum)

Rollups can implement a two-stage block production process: competitive builders bid for the right to construct a block, and a decentralized proposer (often a validator) selects the highest bid. This separates trust and creates a liquid market.\n- Decentralized Censorship Resistance: No single entity controls transaction inclusion.\n- Optimized Execution: Builders compete on gas efficiency and MEV extraction, lowering user costs.

~500ms
Auction Window
-15%
Avg. Fee
03

The Frontier: Encrypted Mempools & SUAVE

Native PBS enables advanced privacy primitives to combat MEV exploitation. Encrypted mempools (e.g., Shutter Network) hide transaction content until after the builder's bid is committed. SUAVE aims to be a universal preference layer for decentralized block building.\n- Fair Ordering: Users are shielded from front-running and sandwich attacks.\n- Cross-Chain MEV: Builders can arbitrage across rollups and L1 in a single block.

0
Visible TXs
Multi-Chain
Scope
04

The Trade-off: Latency vs. Decentralization

Implementing PBS adds ~1-2 seconds of latency to block production due to the auction process. For high-frequency DeFi or gaming rollups, this is a critical design decision. Solutions like pre-confirmations or fast-lane services from builders like BloXroute or Blocknative will emerge to bridge this gap.\n- User Experience: Instant pre-confirms vs. decentralized finality.\n- Market Segmentation: Different rollups will optimize for different use cases.

+1.5s
Added Latency
10x
More Builders
05

The Builder Stack: New Infrastructure Vertical

PBS creates a dedicated market for rollup-specific builder software. Expect specialization: Flashbots SUAVE for cross-domain MEV, EigenLayer AVSs for decentralized proposer networks, and RaaS providers (e.g., Conduit, Caldera) offering PBS-as-a-service.\n- Specialized MEV: Builders will develop rollup-specific strategies (e.g., L2->L1 arbitrage).\n- Rapid Deployment: Rollup teams can outsource PBS complexity to infrastructure providers.

New $B
Market
5+
Major Players
06

The Endgame: Shared Sequencing & Interop

The ultimate evolution is a shared sequencer (e.g., Espresso, Astria) that provides PBS and atomic cross-rollup composability. This moves the PBS auction and MEV market above individual rollups, creating a unified liquidity layer.\n- Atomic Composability: Transactions across multiple rollups settled in one atomic bundle.\n- Network Effects: MEV revenue and security are shared across the ecosystem.

Unified
Liquidity
Atomic
Cross-Rollup
ENQUIRY

Get In Touch
today.

Our experts will offer a free quote and a 30min call to discuss your project.

NDA Protected
24h Response
Directly to Engineering Team
10+
Protocols Shipped
$20M+
TVL Overall
NDA Protected Directly to Engineering Team