Execution is a commodity. The core function of processing transactions—adding numbers, moving tokens—is computationally trivial. The value is in the network's state and user base, not the raw compute.
the-modular-blockchain-thesis-explained
Blog
Why Execution Layer Competition Drives Fees to Zero
The modular blockchain thesis commoditizes execution. With rollup deployment now a solved problem via stacks like Arbitrum Orbit and OP Stack, we analyze the inevitable economic pressure driving transaction fees to the marginal cost of the underlying data layer.
introduction
THE COMPETITIVE FORCE
Introduction: The End of the Execution Monopoly
Block execution is becoming a commodity, driving transaction fees toward zero through relentless competition.
Monopoly rents are unsustainable. Ethereum's L1 and early L2s captured value via execution exclusivity. This invites competition from Arbitrum, Optimism, and zkSync, which replicate state at lower cost.
The endpoint is zero-fee execution. As L2s compete on cost, the price for pure execution converges to the marginal cost of compute, which is effectively zero. Users pay for data, not calculation.
Evidence: Arbitrum One's average transaction fee is $0.10, while Base and Blast subsidize gas to attract users. This is a race to the bottom for the execution layer's profit margin.
key-trends
EXECUTION LAYER COMPETITION
Key Trends: The Commoditization Flywheel
The modular stack separates execution from consensus, creating a hyper-competitive market for block space that relentlessly drives fees toward zero.
01
The Problem: Monolithic Rent Extraction
Ethereum L1 bundles consensus, data, and execution, creating a single-point fee market. This leads to predictable congestion and high, volatile gas fees during demand spikes, acting as a tax on all applications.
- Single Revenue Pool: All dApps compete for the same scarce block space.
- Inelastic Supply: Throughput is capped by the monolithic architecture.
- Value Capture: Fees accrue to the base layer, not to specialized providers.
$100M+
Monthly Fees
~15 TPS
Base Throughput
02
The Solution: Modular Execution Markets
Rollups and L2s commoditize execution by outsourcing it to competitive networks. This creates a race to the bottom on fees while allowing for specialized performance.
- Parallel Blockchains: Rollups like Arbitrum, Optimism, and zkSync compete on cost and speed.
- Shared Security: They inherit Ethereum's security, lowering their own trust assumptions.
- Specialized VMs: Execution environments like FuelVM and Solana SVM (via Eclipse) optimize for specific use cases.
<$0.01
Avg. L2 TX Cost
1000+ TPS
Scaled Capacity
03
The Flywheel: Shared Sequencers & Prover Markets
The commoditization deepens. Shared sequencer networks like Astria and Espresso decouple ordering from execution, creating a market for block building. Similarly, proof aggregation services commoditize ZK-proof generation.
- Sequencer Competition: Execution layers bid for inclusion in shared, decentralized sequencing sets.
- Prover Auctions: Networks like RiscZero and Succinct compete to generate ZK proofs cheapest and fastest.
- Ultimate Outcome: Execution becomes a low-margin, high-volume utility, akin to cloud computing.
~500ms
Finality Target
-90%
Sequencing Cost
04
The Endgame: Intent-Based Abstraction
Users no longer interact with chains. Protocols like UniswapX, CowSwap, and Across use solvers to find optimal execution across the fragmented landscape. The execution layer becomes an invisible commodity.
- User Declares Goal: "Swap X for Y at best price."
- Solver Competition: A network of solvers (PropellerHeads, Revert) competes to fulfill the intent across all L2s and liquidity venues.
- Fee Compression: Solvers drive execution costs to their absolute marginal cost, passing savings to users.
5-10x
Better Price Execution
$0
User Gas Management
deep-dive
THE MARGIN
Deep Dive: The Economics of Perfect Competition
The commoditization of block space forces execution layer profits toward zero, creating a new battleground for value capture.
Execution is a commodity. The fundamental service of ordering and processing transactions is not proprietary; any chain or rollup with an EVM can provide it. This lack of differentiation triggers a race to the bottom on fees.
Validators compete on cost. Profit margins are determined by the delta between the fee users pay and the validator's operational cost. As competition intensifies, this margin compresses toward the cost of capital and hardware.
Layer 2s exemplify this trend. Networks like Arbitrum and Optimism have driven transaction fees below $0.01 by optimizing their sequencer operations, demonstrating that high throughput at near-zero cost is the new baseline.
The value shifts upstream. With execution margins gone, the real economic moat moves to the application layer and settlement guarantees. This is why EigenLayer and restaking protocols are attracting billions—they monetize security, not computation.
THE RACE TO THE BOTTOM
Data Highlight: The Cost Basis of Execution
Comparing the marginal cost structure of leading execution environments. Lower marginal cost enables sustainable fee compression.
| Cost Component | Generalized EVM L1 (e.g., Ethereum) | Generalized L2 (e.g., Arbitrum, Base) | App-Specific Chain (e.g., dYdX, Aevo) | Sovereign Rollup (e.g., Celestia-based) |
|---|---|---|---|---|
Block Production Cost | $100s - $1000s (PoW/PoS consensus) | $10s - $100s (L1 Data + Prover Cost) | < $10 (Tailored VM, no contention) | ~$0 (Sovereign, pays only for data) |
State Growth Cost | High (Global state bloat) | High (Inherits L1 + own state) | Low (Isolated, prunable state) | None (Settlement layer's problem) |
Execution Overhead | 100% (Baseline) | ~80-90% (Optimistic) / ~60-70% (ZK) of L1 | ~20-50% (Custom opcodes, no unused logic) | ~10-30% (Minimal VM, pure execution) |
MEV Capture & Redistribution | Validator Extractable Value | Sequencer Extractable Value (partially redistributable) | Protocol Extractable Value (fully capturable) | Sovereign Extractable Value (fully capturable) |
Fee Market Contention | Global (All apps compete) | Semi-Global (L2 app competition) | None (Single app, predictable pricing) | None (Sovereign, predictable pricing) |
Marginal Cost per TX (Est.) | $0.10 - $2.00+ | $0.01 - $0.10 | < $0.001 | < $0.0001 |
Sustainable Fee Floor | High (Tied to L1 security) | Medium (Tied to L1 data cost) | Very Low (Cost-optimized stack) | Near-Zero (Data-only base layer) |
counter-argument
THE COMMODITY TRAP
Counter-Argument: Can Rollups Differentiate?
Execution layer competition will commoditize rollups, driving transaction fees toward the marginal cost of computation and data availability.
Execution is a commodity. Rollups compete on identical EVM bytecode execution. The only technical differentiators are sequencer latency and proving cost, which are marginal and will be arbitraged away by users and builders.
The real moat is elsewhere. Sustainable differentiation exists in the application layer (e.g., dYdX's orderbook) or the shared sequencer network (e.g., Espresso, Astria). The execution engine itself is infrastructure.
Evidence from L1 history. Ethereum's own history proves this: clients like Geth and Erigon compete on performance, but node operators choose the cheapest, most reliable option. Rollup clients (OP Stack, Arbitrum Nitro) will follow the same path.
Fee compression is inevitable. With multiple ZK-rollup and Optimistic rollup stacks offering near-identical performance, fees will converge to the cost of data availability on Ethereum or an alternative DA layer like Celestia or EigenDA, plus a tiny profit margin.
takeaways
EXECUTION LAYER ECONOMICS
Key Takeaways for Builders & Investors
The commoditization of block space is inevitable; the real value accrues to the settlement and data availability layers.
01
The Problem: The MEV-Abundant Commodity
General-purpose EVM execution is a race to the bottom. The only sustainable revenue for a standalone L1 or L2 is MEV extraction and sequencer rent, which are zero-sum games that degrade UX.\n- Result: Fees trend to marginal cost of compute, which is near-zero.\n- Evidence: L2s like Arbitrum and Optimism already subsidize fees, while Solana's ~$0.001 average fee sets the benchmark.
~$0.001
Fee Benchmark
>90%
Subsidized
02
The Solution: Specialized Execution Environments
Avoid competing on raw TPS. Build execution layers optimized for specific applications where performance is non-fungible.\n- Parallel VMs: Solana, Monad, and Sui win by saturating hardware, making latency the premium.\n- Privacy VMs: Aztec and Aleo monetize data opacity.\n- Intent-Based: UniswapX and CowSwap abstract execution, capturing value at the solver layer.
10-100x
Latency Edge
Niche Fee
Premium
03
The Investment Thesis: Own the Base Layer
Value consolidates where competition is impossible: cryptoeconomic security and data availability.\n- Settlement: Ethereum L1 and Bitcoin are the only truly decentralized global settlement layers.\n- DA: Celestia, EigenDA, and Avail monetize data bandwidth, not computation.\n- Build Here: Applications built on these bases are portable and benefit from execution layer competition.
$10B+
Security Budget
~$0.0015
DA Cost/Blob
04
The Arbitrage: Vertical Integration
Capture value by controlling the full stack from user to settlement. dYdX v4 (own Cosmos chain) and Aevo (own L2) escape shared sequencer economics.\n- Control: Own the sequencer, capture the MEV, and customize the fee market.\n- Risk: You now bear the full security and liveness cost. This only works for apps with >$100M+ TVL and >10k daily active users.
100%
MEV Capture
$100M+
TVL Threshold
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