Sequencer revenue and L2 economics

L2 sequencers collect fees from users and pay costs to Ethereum. The difference is revenue. Simple in theory. Complex in practice because L1 gas prices fluctuate wildly and sequencer centralization creates capture questions.

Key takeaways

Sequencers earn revenue from user fees but must pay L1 for data availability and settlement
Operating costs fluctuate with L1 gas prices, creating volatile margins even when user fees are stable
Centralized sequencers capture 100% of revenue; decentralized sequencers split it among operators
Sustainability depends on user fees exceeding L1 costs consistently across gas price regimes
Competitive dynamics between L2s will compress fees, potentially eliminating profits

What a Sequencer Does

Sequencers order transactions and batch them for submission to Ethereum L1. Users send transactions to the L2. The sequencer determines ordering, executes state transitions, and posts data to L1 for security. This service creates the L2 user experience.

The sequencer role is powerful. It controls ordering, which creates MEV opportunities. It determines which transactions are included. It batches efficiently or inefficiently, affecting costs. Whoever controls the sequencer captures these economic rents.

Where Sequencer Revenue Comes From

Users pay transaction fees to the L2 for inclusion and execution. These fees are the gross revenue line. An L2 collecting $100M annually in user fees has $100M gross revenue.

The sequencer operator receives these fees initially. What happens next depends on governance and architecture. Centralized sequencers (one entity) keep everything. Decentralized sequencers split revenue among operators. Some L2s burn fees or distribute them to tokenholders.

Operating Costs and Constraints

L1 data availability is the primary cost. Every L2 transaction must be posted to Ethereum for security. This posting costs L1 gas. When L1 gas is 20 gwei, costs are modest. At 200 gwei, costs are 10x higher.

Calldata costs dominate for rollups using Ethereum's calldata. EIP-4844 (blobs) reduced these costs significantly. L2s migrating to blobs saw cost reductions of 90%+. This massively improved margins but also enabled fee competition.

Settlement costs are secondary. Some L2s pay for L1 state updates and proof verification. These costs are more predictable than data availability but still tied to L1 gas prices.

Infrastructure costs are operational. Running nodes, maintaining software, providing RPC access. These are relatively fixed and small compared to L1 costs. A well-run sequencer might spend $1-5M annually on infrastructure for a network handling billions in volume.

Centralized vs Decentralized Sequencers

Most L2s currently use centralized sequencers. One entity (often the founding team) controls transaction ordering. They capture 100% of sequencer revenue. This creates massive income: Optimism and Arbitrum sequencers have generated hundreds of millions in profit.

Centralized capture creates awkward tokenomics. The sequencer earns millions. Tokenholders receive nothing unless the entity voluntarily shares. Some L2s promise future decentralization and revenue sharing. Promises aren't mechanisms.

Decentralized sequencers split revenue among multiple operators. MEV capture is more complex. Revenue distribution requires coordination. The technology is less mature. Few L2s have achieved this successfully.

The transition from centralized to decentralized sequencing is economically significant. Current centralized operators earn all revenue. Decentralization redistributes that revenue. If not carefully designed, it could reduce total value capture or create new rent extraction.

Sustainability of L2 Revenue

Sustainability requires user fees to exceed L1 costs across market cycles. During cheap L1 gas, this is easy. L2s earn 80-90% margins. During expensive L1 gas, margins compress to 20% or go negative.

Some L2s subsidize users during high gas periods. They charge low fees even when L1 costs spike. This creates user loyalty but burns capital. It's only sustainable if funded by treasuries or investors.

Long-term sustainability also requires defending against competition. As more L2s launch, they compete on fees. Users are price-sensitive. This drives fees toward marginal cost (L1 posting). Profit margins compress industry-wide.

Competitive Dynamics Between L2s

L2s compete primarily on fees and performance. Lower fees attract users. Faster confirmation attracts applications. This creates a race to the bottom on pricing.

Network effects provide some moat. Applications deployed on Arbitrum stay there due to integration costs. Users with assets bridged prefer staying on-chain. But network effects are weaker than on L1. Liquidity fragments easily.

The endgame might be minimal sequencer revenue across all L2s. Fees compress to barely above L1 costs. Profits disappear. L2s become infrastructure maintained by foundations, not profit-generating businesses. This is good for users, bad for sequencer operators and tokenholders.

See live data

Links open DefiLlama or other external sources.

Related Concepts

Protocol revenue: Understanding what L2s retain after costs
Fees vs revenue vs profit: Why L2 fees aren't pure profit
Ethereum L2 fees: Detailed breakdown of L2 fee structure
Ethereum gas fees: L1 costs that L2s must pay
Onchain cash flow: Tracking L2 profitability through stablecoin flows
Cost of funds in DeFi: L1 costs function as cost of goods sold
Ethereum: The L1 that L2s settle to

FAQ

Do L2 tokens capture sequencer revenue?

Usually no. Most L2s have centralized sequencers controlled by entities separate from token governance. The sequencer revenue goes to those entities, not tokenholders. Some L2s promise future revenue sharing, but mechanisms aren't implemented. Check actual distributions, not promises.

Are L2s profitable?

Many are, especially during low L1 gas periods. But profitability is volatile and depends on L1 costs. Some L2s subsidize fees, operating at losses to gain market share. Post-EIP-4844, margins improved significantly. Long-term profitability depends on sustaining fees above marginal costs.

Why don't L2s just charge higher fees?

Competition. If Arbitrum charges 10x higher fees than Optimism, users and developers migrate. Network effects are real but not insurmountable. L2s are forced to price near competitors or justify premium through performance, security, or ecosystem advantages.

What happens when L1 gas spikes?

L2 costs spike proportionally. Some L2s maintain low user fees and absorb losses. Others raise fees to maintain margins. Users might experience worse UX or higher costs. Long spike periods can burn through treasury reserves if L2s subsidize operations.

Can sequencers capture MEV?

Yes. Centralized sequencers can extract MEV from transaction ordering. This is additional revenue beyond user fees. Some estimate this could be significant. Decentralized sequencers require mechanisms to split MEV fairly, adding complexity.

Will L2 sequencer revenue disappear?

Possibly. Intense competition could drive fees to marginal cost. If users are price-sensitive and L2s are substitutable, profit margins compress to near-zero. The infrastructure persists but becomes a nonprofit public good rather than a profit-generating business.

How do blobs affect sequencer economics?

Blobs reduced L1 posting costs by 90%+. This massively improved margins. L2s went from breakeven to highly profitable. But it also enabled lower user fees, intensifying competition. The cost savings benefited users as much as sequencers.

Cite this definition

Sequencer revenue is L2 transaction fees collected from users minus L1 data availability and settlement costs, creating volatile gross margins dependent on Ethereum gas prices, with value capture determined by sequencer centralization and whether mechanisms exist to distribute profits to tokenholders.