Products at the L1-L2 Boundary: A First-Principles Analysis of What Justifies Synchronous Composability

tl;dr

• Most economic activity at chain borders does not require synchronous atomicity. Of the five major border activities – arbitrage, settlement, insurance/hedging, cross-domain lending, and fee optimization – only two (atomic arbitrage and cross-chain flash loans) have a hard requirement for synchronous execution. The rest can be served with async intents and fast confirmations.
• The efficiency gap between inventory-based and atomic arbitrage is real but narrow. Inventory-based arbitrageurs already capture two-thirds of cross-chain arb volume with 9-second median settlement. Synchronous composability would eliminate ~$2-4M annually in leg risk losses and reduce capital lockup – meaningful for searchers, but not a market-defining improvement.
• Cross-chain flash loans are the only genuinely novel product class enabled by synchronous composability. The CRATE protocol (February 2025, academic) demonstrates feasibility with 0.32-0.75x gas overhead relative to standard ZK rollup transactions. But flash loans are an efficiency tool for sophisticated actors, not a mass-market product.
• “Unified AMM liquidity” across rollups is likely worse than concentrated liquidity on one chain. Synchronous access to fragmented pools introduces proving latency, coordination costs, and MEV surface area that offset the theoretical depth advantages. Aerodrome’s success on Base alone is the empirical counterpoint.
• The real product opportunity is cross-domain collateral – lending against assets on other chains without bridging. This would unlock capital efficiency at scale. But it can be built with optimistic verification and fast finality rather than requiring synchronous atomicity.
• Synchronous composability increases total MEV while concentrating it. Cross-domain atomic execution creates new extraction opportunities (cross-rollup sandwich attacks, unified liquidation cascades) that benefit sophisticated searchers. MEV redistribution mechanisms remain unsolved.
• An L1-L2 “trade agreement” requires explicit terms on six dimensions – fee sharing, MEV allocation, liquidity commitments, exit terms, dispute resolution, and standards compliance – none of which the EEZ has specified. Without these, the EEZ is a technical framework, not an economic compact.

Table of Contents

1. The Border Economics Framework
2. Activity-by-Activity: Does It Require Synchronous Atomicity?
3. Concrete Product Designs Requiring Cross-Chain Atomicity
4. The MEV Question
5. Designing an L1-L2 Trade Agreement from First Principles
6. Does the EEZ’s Design Constrain or Enable These Economics?
7. Synthesis: The Product Hierarchy

1. The Border Economics Framework

The analogy to international trade is instructive – but it also reveals where it breaks down. At a national border, the following economic activities emerge naturally:

The critical insight from trade economics: borders create friction, and friction creates economic activity. Customs brokers, currency exchangers, trade financiers, and logistics coordinators all exist because borders exist. Removing the border does not eliminate these activities – it eliminates the jobs of the intermediaries who service them and redirects value to whoever controls the unified zone.

This is the first-principles tension at the core of the EEZ: synchronous composability removes the border. Removing the border is good for users but eliminates the economic niche occupied by bridges, cross-chain arbitrageurs, and relay operators. The question is whether the products enabled by a borderless zone create more value than the products destroyed.

2. Activity-by-Activity: Does It Require Synchronous Atomicity?

2.1 Arbitrage: Inventory-Based (Async) vs. Atomic (Sync)

The current state. Cross-chain arbitrage already functions without synchronous composability. Academic research (arXiv:2501.17335, accepted at ACM SIGMETRICS 2025) documented 242,535 cross-chain arbitrage transactions generating $868.64M in trading volume and $8.65M in net profit over September 2023 - August 2024.

Two execution methods dominate:

First-principles analysis. Inventory-based arbitrage is already near-synchronous (9 seconds). The efficiency gain from true atomicity is marginal for this class:

• Capital efficiency improvement: Inventory-based arbs lock capital on every chain they operate across. With N chains, they need N copies of their working capital. Synchronous atomicity reduces this to 1x. For a searcher with $10M across 5 chains, this frees $40M – meaningful at the individual level but small in aggregate market terms.
• Leg risk elimination for bridge-based arbs: The 33% of arbs using bridges face 242-second settlement and real execution risk. Synchronous atomicity would collapse this to zero, converting bridge-based arbs into atomic arbs. This is a genuine improvement, but it benefits a small class of actors.
• Tighter spreads: With lower capital requirements and zero leg risk, arbitrageurs can profitably close smaller price discrepancies. This improves market efficiency – prices across chains stay tighter. The beneficiary is every trader who gets better execution, though the effect is diffuse and hard to measure.

Verdict: Synchronous composability improves arbitrage efficiency but is not required. The market already functions with async arbitrage. The improvement is incremental for existing actors and does not enable a categorically new product. Sync composability is a nice-to-have for arbitrage, not a must-have.

2.2 Settlement: Does T+0 Enable New Products vs. T+7min?

The current state. L2-to-L1 settlement currently takes:

• Optimistic rollups: 7 days (challenge period)
• ZK rollups: minutes to hours (proving + L1 inclusion)
• With fast finality bridges (e.g., Across): ~seconds for soft confirmation, minutes for hard settlement

First-principles analysis. T+0 cross-domain settlement (synchronous) vs. T+minutes (fast ZK) vs. T+hours (current ZK) vs. T+7 days (optimistic):

The critical question is whether the gap between T+minutes and T+0 enables products that justify the infrastructure cost. The products uniquely enabled by T+0 are:

1. Cross-chain flash loans (analyzed below)
2. Same-slot L1-L2 contract calls (smart contract on L2 calling L1 contract and getting result in the same transaction)
3. Atomic multi-leg DeFi transactions (borrow on L1, swap on L2, repay on L1 – all in one tx)

These are real but narrow. The vast majority of DeFi activity – lending, LPing, trading, governance – functions perfectly well with T+minutes settlement. The products uniquely requiring T+0 are all sophisticate-facing (searchers, MEV actors, complex DeFi users), not retail-facing.

Verdict: T+minutes (fast ZK proving) captures 95%+ of the value. The incremental value of T+0 is real but concentrated among sophisticated actors.

2.3 Insurance/Hedging: Managing Cross-Domain Risk

First-principles analysis. Cross-domain risk exists today because bridging is risky. Bridge exploits have cost billions (Ronin: $625M, Wormhole: $320M, Nomad: $190M). Insurance products against bridge risk (Nexus Mutual policies, Risk Harbor) exist but are small.

Synchronous composability eliminates bridge risk rather than insuring against it. If L2 contracts can call L1 contracts directly (as the EEZ promises), the need for bridge insurance disappears – but so does the insurance market.

The hedging use case that remains: cross-domain position risk. If you hold collateral on L2 and a loan on L1, you face the risk that your L2 collateral drops in value but the L1 liquidation cannot execute quickly enough to protect you. Synchronous composability allows the L1 lending protocol to atomically access L2 collateral state, reducing the lag between price movement and liquidation.

Verdict: Synchronous composability reduces the need for cross-domain insurance products rather than enabling them. It improves liquidation efficiency for lending protocols – a meaningful but behind-the-scenes improvement.

2.4 Lending Against Cross-Domain Collateral

First-principles analysis. This is arguably the most commercially significant product category at the L1-L2 boundary. Today, if you hold $1M in stETH on Ethereum L1, you cannot use it as collateral to borrow on Base without first bridging it. This is:

• Slow (bridge latency)
• Risky (bridge exploit risk)
• Capital-inefficient (your stETH is idle during transit)
• Costly (bridge fees + gas on both chains)

A cross-domain lending protocol would allow you to deposit collateral on Chain A and borrow on Chain B, with the lending protocol verifying collateral across chains in real time.

Does this require synchronous atomicity? Here is where the analysis gets interesting:

• Loan origination: Does NOT require synchronous atomicity. You can verify collateral state with a ZK proof of the L2’s state root, and the L1 lending contract can accept this proof asynchronously. This is how protocols like Polymer and Lagrange approach cross-chain state verification. Latency of minutes is acceptable for loan origination.

• Liquidation: This is where the sync/async question becomes sharp. If the borrower’s collateral drops below the liquidation threshold, the lending protocol needs to:
  1. Verify the current collateral value (on Chain A)
  2. Seize the collateral (on Chain A)
  3. Clear the debt (on Chain B)

  With synchronous composability, this is one atomic transaction. Without it, there is a gap between price observation and liquidation execution, during which the collateral can drop further. The lending protocol must set a higher collateral ratio to compensate – meaning less capital efficiency for borrowers.

• The quantitative question: How much does the async liquidation lag cost? If ZK proofs provide state verification in minutes, the additional buffer required is modest (perhaps 2-5% higher collateral ratio). If it is 7 days (optimistic rollups), the buffer is massive and the product is unviable. The improvement from T+minutes to T+0 saves borrowers perhaps 1-3% in collateral ratio – significant at scale but not a binary enabler.

Verdict: Cross-domain lending is the highest-value product at the L1-L2 boundary. Synchronous composability makes it more capital-efficient, but it can be built with fast async verification (T+minutes). The improvement from sync over fast-async is quantitative, not qualitative – better collateral ratios, not a new product category.

2.5 Fee/Cost Optimization (Routing Through Cheapest Path)

First-principles analysis. Users and protocols already route transactions through the cheapest execution path. Intent-based systems (ERC-7683, Across, UniswapX) do this asynchronously – the user expresses an intent, solvers compete to fill it via the cheapest route.

Synchronous composability does not meaningfully improve routing optimization. A solver that can atomically route through L1 and L2 in one transaction has a marginal speed advantage over a solver that routes asynchronously, but the cost savings come from the route selection, not the execution mode.

Verdict: Async intents already handle fee optimization well. Synchronous composability adds negligible value here.

3. Concrete Product Designs Requiring Cross-Chain Atomicity

3.1 Cross-Chain Flash Loans

What exists. The CRATE protocol (February 2025, academic paper from arXiv:2502.04659) is the first rigorous protocol for cross-rollup atomic transaction execution. It enables:

• Cross-rollup flash loans: borrow on Rollup A, use on Rollup B, repay on Rollup A – all atomically
• Achieves finality in 4 rounds on L1
• Gas overhead of 0.64-1.5x compared to conventional ZK rollup transactions
• Uses an off-chain Executor with an extended EVM (XEVM) and General System Contracts for trigger-action synchronization
• Two-phase commit protocol among L1-based Validator smart contracts

First-principles analysis. Cross-chain flash loans are genuinely novel. On a single chain, flash loans enable:

• Capital-free arbitrage
• Collateral swaps (refinancing positions without temporary undercollateralization)
• Self-liquidation (paying off debt with collateral in one transaction)

Cross-chain flash loans extend this to multi-chain DeFi. The use case: borrow USDC on Arbitrum (where rates are lower), use it to refinance a position on Base (where you have collateral), repay – all atomically. Or: flash-borrow ETH on L1, use it as collateral on L2 to extract a profit from an arb, repay on L1.

The honest assessment. Flash loans are powerful but niche. On Ethereum L1, flash loan volume is dominated by a small number of sophisticated actors (primarily arbitrageurs and liquidation bots). Aave’s flash loans generated ~$7B in volume in 2024 but this is a tiny fraction of total DeFi volume and the fees are minimal. Cross-chain flash loans would be even more niche – useful for cross-chain arb and cross-chain liquidation, but not a mass-market product.

Verdict: Cross-chain flash loans are the purest example of a product that requires synchronous atomicity. They are real, technically feasible (CRATE demonstrates this), but commercially niche.

3.2 Unified AMM Liquidity Across Rollups

The thesis. Instead of Uniswap deploying independent pools on Ethereum, Arbitrum, Base, Optimism, etc., a unified AMM would pool all liquidity into a single logical pool accessible from any chain. A trader on Base would access the same ETH/USDC liquidity as a trader on Arbitrum.

First-principles analysis. This sounds compelling but has deep problems:

Arguments for unified liquidity:

• Deeper pools = less slippage for large trades
• LPs earn fees from all chains, improving capital efficiency
• Eliminates the need for LPs to choose which chain to deploy on

Arguments against unified liquidity (analysis, not sourced claims):

1. Concentrated liquidity already solves depth. Uniswap v3/v4 concentrated liquidity allows LPs to achieve capital efficiency 4000x better than v2-style full-range positions. A $10M concentrated position on a single chain can provide execution quality comparable to a $40B full-range position. The “fragmented liquidity” problem is significantly mitigated by concentrated liquidity – the depth problem is solved by better LP strategies, not by pooling across chains.

2. Proving latency degrades execution. If a trade on Base needs to atomically access liquidity state on L1, there is irreducible latency for the ZK proof verification. Even “real-time” proving (the EEZ’s claim) involves computational overhead. For a competitive AMM, any latency disadvantage vs. a native single-chain AMM is a deal-breaker. Searchers and arbitrageurs will always prefer the faster venue.

3. MEV surface area expands. A unified pool accessible from multiple chains creates new MEV vectors: a searcher who sees a pending trade on Rollup A can front-run it by submitting on Rollup B (if both access the same pool). This is worse than single-chain MEV because the attack surface multiplies with each connected chain.

4. Aerodrome is the empirical counterpoint. Aerodrome captures 50-63% of all Base DEX volume with a single-chain deployment. It did not need cross-chain liquidity to achieve dominant market share. Its ve(3,3) flywheel generates sticky, deep liquidity through token incentive design, not through pooling liquidity from other chains. When a protocol can achieve dominance on one chain through mechanism design, the case for unified cross-chain liquidity weakens considerably.

Verdict: Unified AMM liquidity is likely worse than concentrated liquidity on a dominant chain. The theoretical depth advantage is offset by proving latency, expanded MEV surface, and the empirical evidence that single-chain protocols win through mechanism design, not liquidity aggregation.

3.3 Cross-Rollup Lending Protocol

What would this look like? A lending protocol that:

• Accepts collateral on any connected chain
• Allows borrowing on any connected chain
• Manages liquidations across chains atomically
• Maintains a unified risk model across all deployments

First-principles analysis. Aave currently deploys independently on 12+ chains. Each deployment has:

• Its own liquidity pool
• Its own risk parameters
• Its own liquidation market
• No awareness of positions on other deployments

A cross-rollup lending protocol would unify these into one logical system. The benefits are real:

• Unified collateral. A user with $100K in stETH on L1 and $50K in USDC on Arbitrum could borrow $120K on Base against their combined $150K collateral. Today, they can only borrow against each position independently.
• Cross-chain liquidation efficiency. A single liquidation bot can atomically liquidate positions across chains, reducing the risk of bad debt from delayed liquidations.
• Unified interest rate markets. Supply and demand for borrowing are pooled, leading to more efficient rate discovery.

The sync vs. async question for lending. Loan origination does not require synchronous atomicity – it can be done with state proofs and minutes of latency. Liquidation benefits from atomicity but can function with fast async (minutes). The one operation that truly requires synchronous atomicity is atomic cross-chain debt restructuring – refinancing a position that spans chains in a single transaction to avoid temporary undercollateralization. This is a real but edge-case need.

Verdict: Cross-rollup lending is the most commercially viable product at the L1-L2 boundary. Most of its functionality can be built with fast async verification. The marginal benefit of synchronous atomicity is primarily in liquidation efficiency and edge-case debt restructuring.

4. The MEV Question

4.1 Who Captures MEV in a Synchronously Composable System?

First-principles analysis. In a single-chain system, MEV flows through a known pipeline:

1. Searchers identify opportunities (arb, liquidations, sandwich attacks)
2. Searchers submit bundles to block builders
3. Block builders construct optimal blocks and bid to proposers
4. Proposers select the highest-bidding builder
5. MEV is split: searcher profit, builder margin, proposer payment

In a synchronously composable multi-chain system, the pipeline changes:

1. Cross-domain searchers identify opportunities that span chains (cross-chain arb, cross-chain liquidation, cross-chain sandwich attacks)
2. Cross-domain bundle builders construct bundles that atomically execute across chains
3. A shared sequencer or unified proposer must order these cross-domain bundles

The critical question: who plays the role of the unified proposer? In the EEZ, this would presumably be the L1 proposer (since the L2 settles synchronously to L1). This means L1 validators capture cross-domain MEV – which is the based rollup thesis in miniature.

Justin Drake’s numbers (cited in DWF Labs research on based rollups): L1 congestion fees are ~3,200 ETH/day, MEV is ~800 ETH/day, and the ratio is moving toward 99:1 in favor of congestion fees. If cross-domain MEV flows to L1 validators, it is a small addition to an already-dominant congestion fee revenue stream.

4.2 Does Synchronous Composability Increase or Decrease Total MEV?

Analysis (not sourced – reasoning from first principles):

It increases total MEV. Here is why:

1. New MEV types emerge. Cross-chain sandwich attacks become possible: a searcher sees a pending swap on Rollup A that will move the price of an asset also traded on Rollup B, and front-runs on Rollup B. This is impossible without synchronous composability.

2. Liquidation cascades amplify. If collateral on Rollup A secures a loan on Rollup B, a price drop triggers liquidations that span chains. The liquidation on Rollup B dumps collateral on Rollup A, further depressing the price, triggering more liquidations. Synchronous composability makes these cascades atomic and faster – more efficient for liquidators, but also more volatile for borrowers.

3. Arbitrage efficiency improves. Tighter spreads mean more arb opportunities are profitable (lower threshold for what constitutes an arbitrageable discrepancy), which increases total arb volume even as per-trade profit shrinks.

Who benefits? Primarily sophisticated actors. Cross-domain MEV requires:

• Monitoring multiple mempools simultaneously
• Constructing atomic bundles across chains
• Significant capital and infrastructure

The research from the composability thesis report bears this out: five addresses execute >50% of all cross-chain arb trades, with one address capturing ~40% post-Dencun. Synchronous composability would further concentrate this activity.

4.3 MEV Redistribution

Current proposals (from search results):

• Shared sequencer revenue sharing: Espresso and Astria propose that cross-domain MEV revenue is partially rebated to rollups that participate in the shared sequencing network. The mechanism: open auctions for rollup block insertion, with revenue distributed across participating rollups’ DAOs.
• Based rollup MEV flow to L1 validators: Based rollups (Taiko, Surge) route MEV to Ethereum L1 validators by using L1 proposers as sequencers. This aligns L2 MEV with L1 security.
• MEV burn: Proposals to burn MEV revenue (analogous to EIP-1559 fee burn) rather than allowing it to be captured by any specific actor. Not implemented anywhere for cross-domain MEV.

The unsolved problem (per the Medium article “Cross-rollup MEV: the unsolved problem of shared sequencing”): Sequencer incentives remain misaligned because each rollup prioritizes local MEV and throughput over cross-rollup fairness. Revenue allocation across rollups sharing a sequencer is unresolved. Rollups want to know that joining a shared sequencer will generate at least as much revenue as running their own – and no current mechanism provides this guarantee.

Verdict: Synchronous composability increases total MEV and concentrates it among sophisticated actors. MEV redistribution mechanisms are proposed but unsolved. This is a significant economic design challenge that the EEZ has not addressed.

5. Designing an L1-L2 Trade Agreement from First Principles

If the EEZ is an economic zone – analogous to the EU single market or a free trade area – it needs explicit terms. Here is what those terms would need to specify, reasoned from first principles:

5.1 Fee Sharing

The question: What percentage of L2 sequencer revenue flows to L1?

Current reality: L2s paid L1 only ~$10M in 2025, down >90% from $113M in 2024 (post-EIP-4844 blob fee collapse). Base earned ~$55M in net profit; almost none flowed to L1.

What the agreement must specify:

• A fee schedule (fixed, percentage, or auction-based) for L2 participation in the EEZ
• Whether fees are paid in ETH (strengthening ETH as money) or in a neutral unit
• Whether fees scale with L2 revenue, L2 transaction count, or L2 proving costs
• A minimum fee floor to prevent free-riding

The tension: L2s joined Ethereum’s ecosystem partly to avoid paying Solana/Binance-level validator fees. Imposing significant fees on EEZ participants recreates the cost structure L2s were designed to escape. But without fees, L1 has no revenue model from L2 activity – the value accrual crisis deepens.

5.2 MEV Allocation

The question: Who gets cross-domain MEV?

Options:

• L1 validators capture all cross-domain MEV (based rollup model)
• Cross-domain MEV is split between L1 and L2s via a predetermined formula
• Cross-domain MEV is auctioned, with proceeds distributed to a shared treasury
• Cross-domain MEV is burned

The tension: L2s that join the EEZ gain synchronous composability but may lose MEV they currently capture with their own sequencers. An L2 sequencer earning $5M/year in MEV will not voluntarily surrender it without compensation. The agreement must make L2s whole – or demonstrate that shared composability generates enough new MEV to offset the loss.

5.3 Liquidity Commitments

The question: Are there minimum TVL requirements to join the EEZ?

First-principles reasoning: An EEZ with a single dominant chain (Base) and twenty zombie chains provides little value. The composability only matters if there is meaningful liquidity on multiple connected chains. But imposing TVL minimums would exclude new chains and entrench incumbents.

Alternative: Instead of TVL minimums, require that EEZ members maintain certain liquidity infrastructure (e.g., deep ETH/USDC pools, functional lending markets). This sets quality standards without capital barriers.

5.4 Exit Terms

The question: What happens if an L2 leaves the EEZ?

This is critical and unaddressed. If an L2 joins the EEZ and users begin relying on synchronous composability (e.g., a lending protocol accepts L2 collateral for L1 loans), exit creates systemic risk. Loans secured by L2 collateral become undercollateralized if the L2 disconnects and the collateral cannot be verified or seized.

What the agreement must specify:

• Notice period for exit (e.g., 6 months minimum)
• Wind-down procedures for cross-chain positions
• Handling of in-flight transactions at the time of exit
• Liability for losses caused by exit (who compensates users with broken cross-chain positions?)

5.5 Dispute Resolution

The question: What if an L2’s state is contested?

In the current system, optimistic rollups have a 7-day challenge window; ZK rollups have proof verification. In the EEZ, if an L2’s state is proven incorrect after it was used in a synchronous cross-chain transaction, the consequences ripple across chains.

What the agreement must specify:

• Whether cross-chain transactions can be reversed if an L2 state proof is later invalidated
• Who bears the economic loss from a fraudulent L2 state (the L2 operator? an insurance fund? the L1?)
• Governance process for dispute adjudication

5.6 Standards Compliance

The question: What technical standards must EEZ members follow?

Minimum requirements (analysis):

• Must use ETH as gas token (ensures ETH demand from EEZ activity)
• Must support standard ERC token interfaces
• Must submit to EEZ-compatible proving infrastructure
• Must use a compatible VM (EVM or EVM-equivalent)
• Must publish state roots to L1 within defined timeframes

The tension: Stricter standards improve interoperability but reduce L2 sovereignty. An L2 that wants to use a custom gas token or non-EVM execution environment is excluded. The EEZ becomes a club with membership criteria, not an open network.

5.7 The Meta-Question

The biggest gap in the EEZ’s current design is not technical – it is economic. The EEZ has not specified any of these terms. It has announced real-time ZK proving, synchronous composability, and a list of founding members. It has not published:

• A fee schedule
• MEV allocation rules
• Liquidity requirements
• Exit procedures
• Dispute resolution mechanisms
• Standards beyond “ETH as gas token”

Without these, the EEZ is a technical demonstration, not an economic zone. The EU single market works not because goods can cross borders (that is the technical layer) but because there is a 70,000-page acquis communautaire specifying the terms under which they cross. The EEZ’s acquis is currently blank.

6. Does the EEZ’s Design Constrain or Enable These Economics?

6.1 What We Know About the EEZ Architecture

Based on public announcements (EthCC Cannes, March 29, 2026) and Gnosis/Zisk communications:

• Real-time ZK proving: Zisk’s ZKVM can prove Ethereum blocks “in real time” (per Jordi Baylina). Specific proving times, hardware requirements, and costs have not been published.
• Synchronous composability: Smart contracts on connected rollups can call L1 contracts with the same guarantees as if they were on Ethereum itself.
• No bridging required: ETH as default gas token; assets on connected rollups are treated as native.
• Founding members: Aave, Titan, Beaver Build, Centrifuge, xStocks.
• Structure: Swiss non-profit, co-funded by the Ethereum Foundation.
• Detailed benchmarks: “Technical details and performance benchmarks will be shared in the coming weeks.”

6.2 Proving Cost Constraints

Analysis (reasoning from what is known about ZK proving economics):

Real-time ZK proving is expensive. Current ZK proving costs for Ethereum-equivalent execution:

• ZK rollups today (zkSync, Scroll, Polygon zkEVM) spend significant operational budgets on proving. Exact costs are proprietary, but estimates from various sources suggest $0.01-0.10+ per transaction for proof generation.
• “Real-time” proving requires either (a) massive parallel proving infrastructure or (b) a fundamentally more efficient prover. Zisk claims the latter.
• The Ethereum 2026 roadmap includes ePBS (Enshrined Proposer-Builder Separation) targeted for the Glamsterdam hardfork, which extends the block pipeline to 6-9 seconds, making single-slot proving more feasible.

Who bears these costs matters enormously:

• If the L2 bears proving costs, it is an operating expense that reduces L2 profitability. L2s already operate at a loss (per the L2 ecosystem report, only Base was profitable in 2025). Adding real-time proving costs could make EEZ membership uneconomical.
• If the L1 bears proving costs (subsidized by the Ethereum Foundation or protocol inflation), it is a public good investment with unclear long-term sustainability.
• If users bear proving costs (via higher gas fees), it makes EEZ chains more expensive than non-EEZ chains, creating a competitive disadvantage.

6.3 What the Architecture Enables

If the proving costs are manageable (a significant “if”), the EEZ architecture enables:

1. Same-slot L1-L2 contract calls – the core primitive for all synchronous composability products
2. Unified state verification – L1 contracts can trustlessly read L2 state within the same slot
3. Atomic cross-chain transactions – transactions that span L1 and connected L2s execute as one

These are powerful primitives. The question this report has been asking is whether the products built on these primitives justify the cost and complexity. The answer, based on the activity-by-activity analysis above, is:

• Strong justification: Cross-domain collateral/lending (large market, real capital efficiency gains)
• Moderate justification: Cross-chain flash loans and atomic arbitrage (real but niche)
• Weak justification: Unified AMM liquidity (likely inferior to concentrated single-chain liquidity)
• No justification: Fee optimization, simple asset transfers (async intents handle these)

7. Synthesis: The Product Hierarchy

Ordering the L1-L2 boundary products by commercial viability and sync-composability dependence:

The Bottom Line

Synchronous composability is a powerful technical primitive whose strongest commercial justification is cross-domain lending – a product that does not strictly require it. The only product that truly requires synchronous atomicity (cross-chain flash loans) has a small addressable market. The products with large addressable markets (lending, arbitrage, routing) can all function with fast asynchronous alternatives.

This does not make the EEZ worthless. It makes it a marginal improvement to existing economic activities rather than an enabler of new ones. The analogy: moving from T+2 to T+1 settlement in equities improved capital efficiency and reduced counterparty risk. It did not create new financial products. Moving to T+0 (instantaneous settlement) would further reduce risk – but nobody is building new financial products that specifically require instantaneous settlement. The improvements are real, quantitative, and invisible to end users.

The EEZ’s real challenge is not technical (they appear to have solved real-time proving). It is economic: defining the terms under which L2s join, value flows, disputes are resolved, and members exit. Without that, it is a highway without toll booths, on-ramps, or traffic laws – technically impressive and economically indeterminate.

Sources

• CRATE: Cross-Rollup Atomic Transaction Execution (arXiv)
• Gnosis and Zisk announce ‘Ethereum Economic Zone’ (The Block)
• Gnosis and Zisk Launch Ethereum Economic Zone (Blockonomi)
• Ethereum Economic Zone Aims to Unify $40 Billion (BigGo Finance)
• Astria: The Shared Sequencer Network
• The Espresso Sequencer (HackMD)
• Based Espresso: ad-hoc shared sequencing (HackMD)
• 10 Cross-Rollup MEV Headaches Coming in 2026 (Medium)
• Cross-chain MEV: Challenges and Solutions (Stanford Blockchain Review)
• Coordination is Power: cross-domain MEV (HackMD)
• Cross-rollup MEV: the unsolved problem of shared sequencing (Medium)
• Revenue allocation in shared sequencing (Flashbots Collective)
• Is the Future of Ethereum Rollups Based? (DWF Labs)
• Based Rollup Economics (Taiko Mirror)
• Defragmenting Liquidity with Asynchronous and Synchronous Composability (Espresso)
• Vitalik: Synchronous atomic composability is overestimated (Bitget News)
• Scaling Ethereum L1 and L2s in 2025 and beyond (Vitalik)
• Ethereum Adopts Zero-Knowledge Proof Validation in 2026 (MEXC)
• Ethereum Will Start Scaling Exponentially With ZK in 2026 (CoinTelegraph)
• SoK: Cross-Domain MEV (arXiv)
• Cross-Chain Arbitrage: The Next Frontier of MEV (ResearchGate)
• Rollup Sequencer Economics: Open Questions (a16z crypto)
• Ethereum builders propose ‘economic zone’ (CoinTelegraph via TradingView)
• New Ethereum project aims to fix network fragmentation (CoinDesk)