In H1 2025, application-layer categories (DeFi, wallets, aggregators) captured roughly 71% of the $9.7B in user-paid onchain fees, while base-layer blockchains took just 22%. Ethereum's share of total onchain fees collapsed from over 40% in 2021 to under 3% by 2025 — even as overall network activity surged dramatically.
The report uses the "Fat Stack" framework: the argument that value capture in 2026 is distributed across three measurable buckets — protocol fees, app retained revenue (take-rate), and MEV rents — and that the decisive advantage tends to sit with whoever controls transaction routing and market design, not necessarily the base layer.
Key findings investors and analysts will want:
Why L1 tokens should be re-evaluated as settlement franchises, not fee machines
How wallets like Phantom are monetising orderflow in a way that rivals base-layer economics
Why MEV is a structurally overlooked third bucket of value capture — and now on regulators' radar
The Solana vs. Ethereum case study: even a monolithic L1 produces a fat app layer, not a fat protocol layer
Three measurable falsifiers that would break the Fat Stack thesis
Read the full report below