The growth of Ethereum layer 2 networks has created an economic segment with revenue, cost structures, and distribution models.
Public debate often focuses on lower fees for end users. The relevant issue for the crypto sector is different. L2s operate as infrastructure businesses. They purchase settlement security from Ethereum and sell execution capacity to applications and users.
The current model is profitable, but it contains an economic disconnect requiring correction. The problem is not an L2 design failure. The problem is a consequence of modular architecture and competition for activity.
Revenue: The Sequencer as Capture Center
The primary revenue source for an L2 is the sequencer. A sequencer receives transactions, defines ordering, executes operations off-chain, and batches results for data submission to Ethereum. Users pay a sequencing fee for access to execution capacity. The fee is lower than layer 1 gas, but the fee remains positive.
The sequencer can also capture MEV through control of inclusion order. The combination of fees and MEV forms the main revenue line for Arbitrum, Optimism, Base, and other networks built on modular stacks.
The sector should recognize a technical fact: the sequencer is not a neutral component. The sequencer is infrastructure with economic power and extraction capacity. Sequencer decentralization remains pending across several networks. Current centralization permits elevated margins.
My opinion is pragmatic. L2 teams should publish sequencer revenue, MEV capture, and cost allocation with the same discipline applied to validator metrics on Ethereum.
Costs: Data Availability and Operations
The dominant cost for an L2 has historically been data publication to Ethereum. Before EIP-4844, L2s compressed transactions and submitted data as calldata to layer 1. Data availability spending represented 80% to 95% of total cost. The introduction of blobs under EIP-4844 changed the structure. Blobs provide data space with independent pricing and lower cost than calldata.
Cost reduction exceeded 90% under normal conditions. After the change, L2 profitability increased substantially. Operating costs include nodes, RPC services, storage, monitoring, bridges, engineering teams, and incentive programs. The sum of operating costs plus Ethereum settlement defines gross margin for each network.
Observed Profitability
Public data from several L2s shows elevated profit margins. Base has reported sequencer margins near 85% during high activity periods. Arbitrum has operated with margins around 55%. Ecosystem treasuries accumulate revenue for development, incentives, and acquisitions.
Comparison with Ethereum is unavoidable. An L2 can generate millions of dollars in daily revenue and pay layer 1 a minor amount for settlement security. My technical opinion is clear. The differential is not an accident. The differential results from modularity. Ethereum sells security and consensus. The L2 sells execution and user experience. The problem appears when execution demand grows and layer 1 security demand does not grow in proportion.
ETH value capture lags. The crypto sector should measure L2 success through net settlement payments to Ethereum, not only through transaction volume or total value locked.
Revenue Sharing: Superchain and Arbitrum
L2 ecosystems have created revenue sharing mechanisms. Optimism Superchain requires chains built on OP Stack to contribute the greater of 2.5% of revenue or 15% of on-chain profit. Base has represented a majority of collective revenue in certain periods.
Arbitrum Expansion Program establishes a revenue share of 10% for qualifying Orbit chains. Of the percentage, 80% goes to the Arbitrum DAO treasury and 20% goes to the Developer Guild. Arbitrum DAO also earns from licensing and priority sequencer access auctions, including Timeboost.
My opinion is clear
Mechanisms are necessary but insufficient. The mechanisms distribute revenue within the L2 ecosystem. The mechanisms do not resolve incentive alignment with Ethereum. A more ambitious model should link a portion of L2 revenue to security payments on layer 1.
L2 success does not automatically translate into value for ETH. L2s pay minimal fees to Ethereum for settlement. ETH burn from base fees on layer 1 decreases when activity migrates to L2. Ethereum security economics depends on layer 1 block space demand and staking. If L2s absorb most activity and do not compensate layer 1, the sustainability of the combined system remains uncertain.
I do not argue L2s should subsidize Ethereum without service. I argue modular architecture needs an explicit compensation mechanism. Security is not a free good. Low-cost data availability is an implicit subsidy from Ethereum to L2. The community should debate whether the subsidy is temporary, conditional, or permanent.
Proposals and Opinion
Several technical and economic paths exist. First, a minimum settlement fee per blob or per aggregated transaction. A minimum fee introduces a revenue floor for Ethereum and reduces reliance on calldata burn. Second, ETH staking as a security requirement in bridges and decentralized sequencers.
If L2s require ETH collateral, demand for ETH increases and shared security strengthens. Third, blob auctions with a reserve price and partial allocation to validators. Fourth, service-level contracts between L2 and Ethereum with penalties for non-compliance. Fifth, competition among L2s for users and developers.
Competition already pressures margins and forces reinvestment in infrastructure. My opinion is clear. The sector should abandon the narrative of L2s as a neutral extension of Ethereum. L2s are infrastructure firms with pricing power and extraction capacity.
Regulation or technical governance can alter the equilibrium. The cleanest solution is economic. The price of security should reflect real cost.
The risk is disconnection between L2 growth and ETH value. The crypto community should demand incentive alignment, transparency in revenue and cost metrics, and compensation mechanisms toward layer 1.
