PBS is now a protocol primitive. The separation of block proposal from construction, pioneered by Ethereum's PBS roadmap, is the blueprint for modular chain design. It enables specialized roles for execution, settlement, and data availability layers like Celestia and EigenDA.
the-modular-blockchain-thesis-explained
Blog
The Future of Proposer-Builder Separation in Modular Chains
The modular blockchain thesis doesn't eliminate MEV; it stratifies it. This analysis argues that rollups will require their own PBS implementations, creating a layered, competitive market for builders across the execution, settlement, and data availability stack.
introduction
THE UNBUNDLING
Introduction
Proposer-Builder Separation is evolving from an L1 scaling tool into the core architectural principle for modular blockchains.
The builder market is fragmenting. Unlike Ethereum's centralized builder dominance, modular chains create a multi-layered builder ecosystem. Rollups like Arbitrum and Optimism now operate as proposers, outsourcing block building to specialized sequencers or shared networks like Espresso and Astria.
This creates new attack vectors. Decoupling across sovereign layers introduces complex trust assumptions. A malicious builder on a rollup can censor or reorder transactions without the L1's knowledge, demanding new cryptographic proofs and slashing mechanisms.
Evidence: The Flashbots SUAVE initiative, aiming to decentralize block building, demonstrates that PBS is no longer just about MEV—it's about interoperable execution environments across a multi-chain landscape.
thesis-statement
THE ARCHITECTURAL IMPERATIVE
Core Thesis: The MEV Stack Replicates Downwards
Proposer-Builder Separation (PBS) is not an Ethereum quirk; it is the inevitable architectural pattern for any modular chain seeking credible neutrality and scale.
PBS is a fractal pattern. The economic and security logic that created specialized builders and proposers on Ethereum L1 replicates at every modular layer. Rollup sequencers are the new proposers, creating a secondary MEV market for L2 block building.
Specialization creates new extractable value. The separation of execution, settlement, and data availability layers introduces new inter-domain arbitrage opportunities. MEV extraction shifts from simple DEX arbitrage to complex cross-rollup and cross-domain strategies.
The builder role fragments. On modular chains, the monolithic builder role splits. Specialized L2 builders (e.g., optimizing Arbitrum blocks) compete with cross-domain aggregators (e.g., leveraging Across or LayerZero) that bundle value across chains.
Evidence: Espresso Systems and Astria are building shared sequencer sets that explicitly replicate Ethereum's PBS model. Their goal is to commoditize L2 block production and prevent centralized sequencers from capturing all value.
key-trends
THE FUTURE OF PROPOSER-BUILDER SEPARATION IN MODULAR CHAINS
Key Trends Driving Layered PBS
As modular blockchains fragment execution and settlement, PBS must evolve into a multi-layered system to manage complexity and capture value.
01
The Problem: MEV is a Multi-Domain Game
In a modular stack, value extraction spans execution, settlement, and data availability layers. A monolithic PBS model fails to capture cross-domain arbitrage and forces builders into inefficient, siloed competition.
- Cross-Rollup Arbitrage: Opportunities exist between Uniswap on Arbitrum and Curve on Base.
- Settlement Layer Censorship: Builders can front-run Celestia data attestations or EigenDA proofs.
- Inefficient Bundling: Without a unified view, builders miss ~15-30% of available cross-domain MEV.
15-30%
MEV Missed
Multi-Domain
Arbitrage
02
The Solution: Hierarchical PBS with a Super-Builder
A layered PBS architecture introduces a top-tier coordinator (Super-Builder) that aggregates blocks from specialized sub-builders at each layer (execution, settlement, DA).
- Unified Optimization: The Super-Builder, akin to Across's solver network, constructs the globally optimal chain of blocks.
- Sub-Builder Specialization: Execution builders compete on Arbitrum, settlement builders on Ethereum, DA builders on Celestia.
- Revenue Sharing: Creates a $100M+ market for block-space derivatives and builder bonds.
$100M+
Market Size
Global Optima
Block Building
03
The Enabler: Intents as the Universal Interface
User intents, popularized by UniswapX and CowSwap, abstract transaction construction. This allows the PBS system to work with high-level user preferences rather than raw calldata.
- Portable Liquidity: An intent to "swap X for Y at best price" can be fulfilled on Optimism, zkSync, or via a LayerZero cross-chain message.
- Builder Competition Shift: Builders compete on fulfillment quality and cost, not just transaction ordering.
- User Experience: Reduces failed tx gas waste by ~90% and abstracts chain-specific complexity.
~90%
Gas Waste Reduced
Chain-Agnostic
Execution
04
The Risk: Centralization of the Coordination Layer
The Super-Builder role becomes a critical central point of failure and censorship. Without careful design, it recreates the miner extractable value (MEV) centralization problem at a higher level.
- Trust Assumption: The coordinator must be credibly neutral; a role Flashbots SUAVE aims to fill.
- Cartel Formation: Top sub-builders could collude, extracting >50% of total MEV.
- Protocol-Enforced PBS: The endgame may require PBS logic baked into the settlement layer protocol itself.
>50%
MEV Extraction Risk
Critical Role
Coordinator
PROTOCOL ARCHITECTURE
The Modular PBS Landscape: Who Builds What?
A comparison of architectural approaches to Proposer-Builder Separation (PBS) for modular execution layers, rollups, and shared sequencers.
| Architectural Feature / Metric | Native PBS (e.g., Ethereum post-4844) | Rollup-Centric PBS (e.g., Espresso, Astria) | Shared Sequencer PBS (e.g., Radius, Madara) |
|---|---|---|---|
Core PBS Abstraction Layer | Consensus Layer (Beacon Chain) | Execution Layer (Rollup) | Sequencer Network |
Builder Market Liquidity Source | Ethereum Validator Set | Rollup Proposer Set | Staked Sequencer Set |
Cross-Domain MEV Capture | Via mev-boost & SUAVE | Native rollup orderflow (e.g., via Espresso) | Enforced via encrypted mempool (e.g., Radius) |
Time to Finality Impact | 12s (Slot Time) | < 2s (Rollup Block Time) | Network Latency Bound |
Builder Censorship Resistance | Yes (crLists, Inclusion Lists) | Varies (Rollup Governance) | Yes (Threshold Encryption) |
Primary Revenue Stream | Ethereum Priority Fees + MEV | Rollup Transaction Fees + MEV | Sequencing Fees + MEV Share |
Key Dependency | Ethereum L1 Finality | Underlying DA Layer Finality | Shared Sequencer Network Liveness |
Example Implementations / R&D | mev-boost, Builder API, SUAVE | Espresso, Astria, AltLayer | Radius, Madara, Fairblock |
deep-dive
THE ARCHITECTURE
Deep Dive: The Mechanics of a Layered Builder Market
Proposer-Builder Separation (PBS) evolves into a multi-layered market where specialized builders compete for execution rights across modular chains.
Specialization defines the layered market. PBS creates distinct roles: proposers (validators) sell block space, builders compete to fill it. This separation forces builders to specialize in MEV extraction, cross-domain atomicity, and local fee optimization to win auctions. The result is a competitive market for block production, not just validation.
Modularity fragments the builder landscape. A monolithic chain like Ethereum has one builder market. A modular stack (e.g., Celestia DA, Arbitrum L2, EigenLayer AVS) creates multiple, interdependent builder markets. Builders must now optimize for specific execution environments, data availability costs, and shared security models, creating niches for rollup-specific builders.
Cross-domain MEV is the killer app. The most valuable builders will be those that capture atomic arbitrage across L2s or between L1 and L2. This requires sophisticated infrastructure like Flashbots SUAVE for intent coordination or Across Protocol's unified liquidity model. Builders become cross-chain system integrators.
Evidence: Ethereum's post-merge block building is dominated by a few entities like Flashbots and bloXroute, proving the economic viability of specialization. The next phase sees AltLayer and Conduit building rollup-specific PBS frameworks, demonstrating the demand for layered solutions.
counter-argument
THE MODULAR DIVIDE
Counter-Argument: Will Shared Sequencing Kill Rollup PBS?
Shared sequencing redefines, rather than eliminates, the need for proposer-builder separation within modular stacks.
Shared sequencers centralize ordering but decentralize block building. Networks like Espresso and Astria provide a canonical transaction order for multiple rollups, but the construction of the actual execution payload remains a separate, competitive market. This creates a new PBS layer between the shared sequencer and individual rollup provers.
The PBS model shifts upward. Instead of a rollup's single sequencer selling blockspace, the shared sequencer's output becomes the auctionable resource. Builders now compete to assemble the most valuable cross-rollup bundles from this ordered stream, optimizing for MEV extraction across chains like Arbitrum and Optimism simultaneously.
This creates a two-tiered auction. The first auction is for the right to sequence (governed by the shared sequencer's consensus). The second is the builder market for execution, where entities like Flashbots and bloXroute compete to produce the most profitable block proofs. This specialization increases efficiency.
Evidence: Espresso's testnet integrates with rollups like Caldera, demonstrating that the sequencer output is a raw feed, not a built block. The economic complexity of cross-rollup MEV necessitates sophisticated builders, preserving the PBS dynamic at a higher layer of abstraction.
risk-analysis
PBS IN MODULAR STACKS
Risk Analysis: The Fragmentation Trade-Offs
Proposer-Builder Separation (PBS) is essential for MEV management and chain neutrality, but its implementation across modular layers introduces systemic complexity.
01
The Problem: Cross-Domain MEV Cartels
Builders can vertically integrate across execution, settlement, and data availability layers, re-centralizing power. A builder controlling a dominant sequencer rollup and a stake in the DA layer can extract and censor transactions across the entire stack, negating PBS's core promise.
>33%
Stake Threat
Multi-Chain
Attack Surface
02
The Solution: Enshrined PBS & Force-Inclusion Lists
Hard-coding PBS at the settlement layer (e.g., Ethereum's ePBS) removes builder trust assumptions. Coupled with force-inclusion lists at the rollup level, this guarantees transaction processing and prevents censorship, even against a malicious builder coalition.
L1-Guaranteed
Neutrality
0-Latency
Censorship Slots
03
The Problem: Liquidity Fragmentation in Builder Pools
Builders require capital for cross-domain arbitrage and bid collateral. In a multi-rollup world, this capital is siloed per chain, reducing efficiency and raising barriers to entry. This leads to higher costs for users and fewer, more powerful builders.
$B+
Capital Locked
<10
Major Builders
04
The Solution: Shared Sequencing & Unified Auctions
A shared sequencer (e.g., Astria, Espresso) aggregates rollup blocks, enabling a single, capital-efficient builder market. Projects like SUAVE aim to create a unified, cross-chain block space auction, decoupling builder intelligence from execution.
90%+
Capital Efficiency
~100ms
Cross-Rollup Latency
05
The Problem: Unverifiable Builder Commitments
In a modular chain, a builder's promise to include a rollup block is just that—a promise. If the builder fails to post data to the DA layer or proof to the settlement layer, the rollup halts. Light clients and bridges cannot independently verify builder reliability.
Hours-Days
Failure Detection
User Funds
At Risk
06
The Solution: ZK Proofs of Publication
Builders must generate a ZK proof of DA data availability (e.g., using Celestia's data availability sampling or EigenDA's proofs) as part of the block bid. This allows for trust-minimized, instant bridging and slashing for non-performance, moving security from social consensus to cryptography.
~1-2s
Verification Time
Cryptographic
Security Guarantee
future-outlook
THE ARCHITECTURAL SHIFT
Future Outlook: The Vertical Integration Play
Proposer-Builder Separation will evolve from a layer-1 concept into a core design pattern for vertically integrated modular stacks.
PBS becomes a stack primitive. The separation of block building from proposing is not just for Ethereum. It is the logical endpoint for any modular chain seeking credible neutrality and MEV management. Rollups like Arbitrum and Optimism will implement native PBS to prevent centralized sequencers from extracting value.
Builders will vertically integrate. Specialized builders like Flashbots SUAVE will not just compete for Ethereum blocks. They will operate dedicated sequencer nodes for rollups, creating optimized execution bundles that span multiple layers. This creates a cross-domain MEV market.
The integration creates new risks. A builder controlling both L1 and L2 sequencing introduces centralization vectors and cross-layer censorship risks. The counterbalance is a robust ecosystem of competing builders and enforceable slashing conditions via protocols like EigenLayer.
Evidence: The design of Celestia's Blobstream and EigenDA as data availability layers presupposes a PBS-like model where rollup sequencers (builders) post data, and a separate set of validators (proposers) attest to its availability. This is PBS for data.
takeaways
PBS IN MODULAR CHAINS
Key Takeaways for Builders and Architects
Proposer-Builder Separation is evolving from an L1 scaling tool into the fundamental coordination layer for modular blockchains.
01
The Problem: PBS is a Single-Chain Tool
Traditional PBS (e.g., Ethereum's MEV-Boost) is designed for monolithic execution. It fails in modular stacks where execution, settlement, and data availability are decoupled. Builders must now coordinate across multiple layers, creating new points of failure and latency.
- Fragmented State: A builder's view is limited to its local chain, missing cross-rollup arbitrage.
- Sequencer Risk: Relying on a rollup's centralized sequencer for inclusion breaks PBS guarantees.
- Inefficient Bundles: Cross-domain bundles are impossible without new infrastructure.
~500ms+
Cross-Domain Latency
0
Native Cross-Chain Bundles
02
The Solution: Shared Sequencing Layers
Networks like Astria, Espresso, and Radius are creating neutral, decentralized sequencer sets that serve multiple rollups. This creates a unified block-building marketplace across an entire ecosystem.
- Unified Mempool: Builders access transactions from hundreds of rollups in one place.
- Atomic Cross-Rollup Bundles: Enable complex DeFi strategies spanning Optimism, Arbitrum, and zkSync in a single block.
- Credible Neutrality: Removes rollup-specific sequencer as a trusted party, restoring PBS security.
100+
Rollups Served
Atomic
Cross-Rollup Execution
03
The Problem: MEV is Now Multi-Domain
In a modular world, the most valuable MEV (e.g., DEX arbitrage, oracle updates) exists between chains and rollups. Isolated PBS auctions on each layer leave this value uncaptured and create systemic risk.
- Inefficient Markets: Separate auctions on Ethereum, Celestia, and each rollup prevent global optimization.
- Extraction Complexity: Builders must win auctions on multiple, asynchronous layers simultaneously.
- New Attack Vectors: Adversarial sequencing on one chain can extract value from dependent chains.
$10B+
Cross-Domain TVL
High
Arb Complexity
04
The Solution: Interchain Block Builders
Next-generation builders like Skip Protocol and Briq are architecting systems that participate in PBS auctions across multiple layers. They use intents and shared sequencing to construct globally optimal, cross-chain block bundles.
- Global Optimization: Algorithms that maximize revenue across Ethereum, Avail, and a rollup set.
- Intent Integration: Incorporate user intents (Ã la UniswapX, CowSwap) for efficient cross-domain settlement.
- Revenue Sharing: New models to share interchain MEV with rollup sequencers and users.
10x+
Bundle Value
Multi-Chain
Auction Participation
05
The Problem: Builder Centralization Threatens Rollup Sovereignty
A dominant interchain builder could become a centralized point of control, dictating transaction ordering across dozens of sovereign rollups. This recreates the miner/extractor centralization problem at a higher, more systemic level.
- Single Point of Failure: A bug or attack in a major builder compromises the entire ecosystem.
- Censorship Power: A builder could blacklist transactions across multiple chains.
- Economic Capture: Rollup revenue is funneled to a single external entity.
>50%
Market Share Risk
Systemic
Censorship Risk
06
The Solution: Enshrined PBS & Verifiable Builders
The endgame is PBS baked into rollup and settlement layer protocols (enshrined). Builders become verifiable state transition provers, with their work checked by the base layer. Projects like EigenLayer's EigenDA and Babylon are pioneering cryptoeconomic security for this.
- Protocol-Level Auctions: PBS is a consensus rule, not an off-protocol marketplace.
- Cryptoeconomic Security: Builders stake on the settlement layer (e.g., Ethereum, Cosmos) and are slashed for misbehavior.
- Prover Networks: Builder output is verified by a decentralized prover network before finalization.
Protocol
Level Guarantees
Slashable
Builder Stake
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall