Modular architecture fragments state. Separating execution (rollups) from consensus (L1) and data availability (Celestia/EigenDA) optimizes for scalability but destroys atomic composability across chains. This creates isolated liquidity pools and fragmented order flow, which is the primary attack surface for MEV.
the-modular-blockchain-thesis-explained
Blog
Why MEV Extraction Cannot Be Solved Without Shared Sequencing
The modular thesis has a critical, unaddressed flaw: it fragments MEV markets. Isolated rollup sequencers turn every L2 into a private, extractive fiefdom. This analysis argues that only a shared sequencing layer can create a global market for MEV, enabling credible redistribution and mitigation strategies.
introduction
THE SEQUENCER BOTTLENECK
Introduction: The Modular Paradox
Modular design isolates execution from consensus, but this decoupling creates a new, critical vulnerability for MEV extraction.
Isolated sequencers are MEV vacuums. Each rollup (Arbitrum, Optimism) operates its own sequencer, creating a sealed-bid auction for block space. This prevents cross-domain MEV capture but maximizes extractable value within each domain, as seen in the persistent sandwich attacks on Arbitrum's DEXs.
The paradox is security versus extraction. You cannot have secure cross-chain intent settlement (via Across or LayerZero) without a shared, neutral sequencing layer. Intent-based systems like UniswapX fail without a global view of liquidity, which only a shared sequencer (like Espresso or Astria) provides.
Evidence: The $1.2B annualized MEV. Over 90% of this value originates from DEX arbitrage and liquidation, operations that become exponentially more profitable and difficult to mitigate when order flow is siloed across dozens of rollup sequencers.
key-trends
WHY SHARED SEQUENCING IS INEVITABLE
The Fragmentation Fallacy: Three Unavoidable Trends
Isolated rollup sequencers create MEV silos, fragmenting liquidity and security. True neutrality requires a shared, competitive layer.
01
The Problem: MEV Silos & Inefficient Markets
Each rollup is a captive market for its sequencer. This fragments liquidity and creates localized, inefficient MEV extraction. A cross-rollup arbitrage opportunity between Arbitrum and Optimism cannot be captured atomically, leaving billions in value inefficiently priced.
- Fragmented Liquidity: Limits arbitrage efficiency and price discovery.
- Sequencer Monopolies: Single points of control extract maximal value from their isolated domain.
- Suboptimal Execution: Users get worse prices as cross-domain opportunities are missed.
$10B+
Fragmented TVL
>20%
Inefficiency Gap
02
The Solution: Cross-Domain Block Space
A shared sequencer like Espresso Systems or Astria creates a unified, competitive marketplace for block space across rollups. This enables atomic cross-rollup bundles, allowing searchers to bid for complex transactions that span multiple chains, capturing true global arbitrage.
- Atomic Composability: Enforce execution across chains or revert all.
- Competitive Auction: MEV revenue is competed away, improving user execution.
- Prover-Agnostic: Works with any rollup stack (OP Stack, Arbitrum Nitro, zkSync).
0ms
Cross-Chain Latency
1 Auction
Unified Market
03
The Trend: Credible Neutrality as Infrastructure
The endgame is a sequencing layer that is credibly neutral, like a public good. This mirrors the evolution from validator extractable value (VEV) to proposer-builder separation (PBS) on Ethereum. Protocols like SUAVE aim to be this neutral layer, decoupling execution from consensus.
- Decoupled Design: Separates block building, ordering, and execution.
- Censorship Resistance: No single entity controls transaction inclusion.
- Economic Security: Staked operators with slashing ensure liveness and correctness.
1000+
Node Operators
Ethereum L1
Security Root
deep-dive
THE DATA
The Inevitable Economics of Isolated MEV
Isolated sequencers create fragmented liquidity and arbitrage opportunities that guarantee MEV extraction will persist.
Isolated sequencers guarantee MEV. A single chain's sequencer can only order transactions within its own state. Cross-domain arbitrage between Ethereum, Arbitrum, and Optimism creates latency races that isolated sequencers cannot coordinate to capture.
Shared sequencing is the only solution. A unified ordering layer like Espresso or Astria enables atomic cross-rollup bundles. This allows the sequencer to internalize inter-domain arbitrage, reducing the extractable value leaked to searcher bots.
Fragmentation creates economic pressure. Without shared sequencing, protocols like UniswapX and Across must build complex intents to simulate atomicity, pushing costs and complexity onto users and applications instead of the infrastructure layer.
Evidence: The MEV-Boost auction on Ethereum L1 proves the value of block space ordering. In an L2 world, this auction simply moves to the cross-rollup domain, where the highest bidder is the arbitrageur bridging the price gap.
WHY ISOLATED SEQUENCERS LEAK VALUE
The Isolation Tax: Comparative MEV Landscape
Comparison of MEV extraction efficiency and user cost across sequencing architectures, demonstrating the economic penalty of isolated block production.
| Critical Dimension | Isolated Rollup (Status Quo) | Shared Sequencer Network (e.g., Espresso, Astria) | Fully Sovereign L1 (e.g., Solana, Monad) |
|---|---|---|---|
Cross-Domain Arbitrage Capture | 0-15% | 85-99% | 95-99% |
User Cost from MEV (Sandwiching) | 0.5-3.0% of tx value | 0.1-0.5% of tx value | < 0.1% of tx value |
Liquidity Fragmentation Cost | 1-5% slippage on bridges | ~0.3% slippage via intents | 0% (native liquidity) |
Proposer-Builder Separation (PBS) Feasibility | |||
Time to Finality for Cross-Domain Settlements | 12-20 minutes (Ethereum L1) | < 1 minute | ~400ms |
Native Support for Intents / RFQs | |||
Required Trust Assumption | Single sequencer operator | Economic security of network (e.g., stake) | Protocol consensus |
protocol-spotlight
WHY MEV EXTRACTION CANNOT BE SOLVED WITHOUT SHARED SEQUENCING
The Shared Sequencing Vanguard
Isolated rollup sequencers create fragmented liquidity and opaque mempools, turning MEV from a nuisance into a systemic threat. Shared sequencing is the only viable coordination layer.
01
The Fragmented Liquidity Trap
Rollups operating as isolated state machines create fragmented liquidity pools. This prevents atomic arbitrage across chains, turning simple trades into multi-step, high-risk operations ripe for exploitation.\n- Inefficient Markets: Price discrepancies persist longer, creating guaranteed profit for searchers.\n- User Cost: Trades require bridging and multiple transactions, increasing fees and slippage.
$10B+
Fragmented TVL
>30%
Arb Spreads
02
The Opaque Mempool Problem
Each rollup's sequencer has a private mempool, creating information asymmetry. Builders with exclusive access can front-run and sandwich user transactions with impunity.\n- Centralized Power: The sequencer becomes a single point of MEV extraction and censorship.\n- No Fairness: Users cannot compete in a sealed-bid environment, guaranteeing losses.
~100ms
Advantage Window
1 Entity
Visibility
03
Espresso & Shared Sequencing
A shared sequencer like Espresso Systems provides a neutral, decentralized mempool for multiple rollups. It enables atomic cross-rollup bundles and fair transaction ordering via a leaderless HotStuff consensus.\n- Atomic Composability: Enables UniswapX-style intents across chains.\n- MEV Redistribution: Proposer-Builder-Separation (PBS) allows for fair auction and redistribution of captured value.
~500ms
Finality
Multi-chain
Atomic Bundles
04
The Economic Inevitability
MEV is a fundamental byproduct of block space. The question isn't elimination, but capture and redistribution. Shared sequencers are the only mechanism with the economic scale and cross-domain visibility to implement effective PBS.\n- Scale for PBS: Requires high, consistent block revenue to sustain a competitive builder market.\n- Credible Neutrality: Decentralized sequencing prevents the sequencer from becoming the ultimate extractor.
$1B+
Annual MEV
Required
For PBS
05
Astria & the Execution Layer
Astria separates sequencing from execution, providing a shared, decentralized sequencer network that rollups can plug into. This commoditizes the sequencing layer, forcing competition on execution and DA.\n- Interoperability by Default: Native cross-rollup messaging and composability.\n- Reduced Overhead: Rollup teams avoid the complexity of running a validator set.
-90%
Dev Complexity
Shared
Security
06
The Final Hurdle: Adoption
The technical solution exists, but adoption requires overcoming rollup vendor lock-in. Major L2s have economic incentives to maintain control of their sequencer and its revenue. The shift will be driven by user demand for better interoperability and MEV protection.\n- Network Effects: Value accrues to the shared sequencer with the most connected rollups.\n- User Choice: Wallets and dApps will route transactions to chains with fairer sequencing.
2-3 Years
Timeline
Top-Down
Pressure
counter-argument
THE COORDINATION FAILURE
Objection: Can't We Just Democratize Individual Sequencers?
Decentralizing single-chain sequencers fails to solve cross-domain MEV, creating a new, fragmented layer of extractable value.
Decentralized single-chain sequencing is a local optimum. Projects like Espresso Systems or Astria decentralize the sequencer for a single rollup, preventing censorship. This solves one problem but creates another: cross-domain MEV now requires coordination between multiple, independent decentralized sequencer sets, which is impossible without a shared communication layer.
Fragmented sequencing creates extractable inefficiencies. Without a shared sequencer, atomic bundles spanning Arbitrum and Optimism require bribing two separate, uncoordinated validator sets. This reintroduces the very rent-seeking and failed coordination that decentralization aimed to solve, but at a higher, inter-chain level. The value leaks into the inter-sequencer network layer.
The market structure becomes adversarial. Independent sequencer sets for zkSync and Base are incentivized to front-run each other's pending cross-domain transactions. This transforms a technical sequencing problem into a game-theoretic prisoner's dilemma, where the dominant strategy is value extraction, not chain interoperability. The result is worse latency and higher costs for users.
Evidence: The Ethereum PBS (Proposer-Builder Separation) model demonstrates that decentralizing block production without a shared coordination mechanism (like a block builder market) centralizes MEV extraction. Applying this model per-rollup replicates the builder centralization problem N times, creating N points of failure instead of one.
takeaways
THE ARCHITECTURAL IMPERATIVE
TL;DR: The Non-Negotiable Path Forward
MEV is a systemic, network-level problem; solving it requires a systemic, network-level solution.
01
The Problem: Isolated Sequencing Creates MEV Silos
Every rollup running its own sequencer is a self-contained MEV extraction zone. This fragmentation guarantees inefficiency and maximal rent extraction.
- Cross-domain arbitrage becomes a multi-step, high-latency game for searchers.
- Users face worst-case slippage as liquidity is balkanized across chains.
- Validators/sequencers capture ~$1B+ annually from this structural flaw.
$1B+
Annual Extractable Value
100+
Fragmented Pools
02
The Solution: Atomic Cross-Chain Execution
A shared sequencer acts as a single, neutral coordinator for transaction ordering across multiple rollups, enabling atomic bundles.
- Eliminates latency arbitrage between L2s, capturing value for users/protocols.
- Enables native intents and CoW-Swap/Across-like functionality at the infrastructure layer.
- Unlocks composable DeFi with ~500ms finality across the entire rollup ecosystem.
Atomic
Cross-Chain Bundles
~500ms
Cross-Domain Latency
03
The Mandate: Credible Neutrality & Enshrined PBS
Decentralization alone is insufficient; the sequencing layer must be credibly neutral with enshrined Proposer-Builder Separation (PBS).
- Prevents censorship and ensures liveness without a single operator.
- Separates block building from proposing, democratizing MEV revenue distribution.
- Creates a verifiable, open marketplace for block space, akin to Ethereum's roadmap but for L2s.
PBS
Enshrined
0
Trusted Operators
04
The Blueprint: Shared Sequencing as L1 for L2s
Treat the shared sequencer set as a new security base layer—a decentralized sequencing 'L1' for the modular stack.
- Provides economic security through staking and slashing, disincentivizing malicious ordering.
- Standardizes the pre-confirmation interface, creating a universal UX layer for all rollups.
- Future-proofs for EigenLayer restaking, Babylon Bitcoin staking, and other cryptoeconomic primitives.
L1-Like
Security Model
Universal
Pre-Confirmation UX
05
The Alternative: Protocol-Level MEV Capture
Without a neutral shared sequencer, MEV will be captured at the application layer, leading to centralized, extractive markets.
- UniswapX and CowSwap become mandatory for fair execution, adding protocol complexity.
- LayerZero and Across act as ad-hoc sequencers for intents, creating new trust dependencies.
- Results in a patchwork of solutions that increase systemic risk and user cost.
Ad-Hoc
Trust Assumptions
+Complexity
Protocol Overhead
06
The Verdict: An Inevitable Consolidation
The economic and security pressures of MEV will force rollup ecosystems to converge on a handful of shared sequencing networks.
- Astria, Espresso, Radius are competing to define this standard.
- The winner will capture $10B+ in secured TVL and become the most critical piece of infra after Ethereum.
- Rollups that remain isolated will be competitively disadvantaged on cost, UX, and security.
$10B+
Future Secured TVL
3-5
Surviving Networks
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