What the fee-generating assets look like on 22 October 2026
A cross-asset look at annualized fee streams and market-cap-to-fee multiples as of 22 October 2026, with notes on what the figures can and cannot support.
The Spread of Market-Cap-to-Fee Multiples
As of 22 October 2026, the market-cap-to-fees ratio across the fourteen assets in this dataset ranges from 3.5x (Uniswap, annualized from the trailing 30 days) to 18,821x (Bitcoin, annualized from the trailing 30 days). That is a spread of roughly four orders of magnitude. A wide spread of this kind tells a reader that fee generation is distributed very unevenly relative to market capitalization across the group; it does not, by itself, tell a reader anything about whether any individual multiple is appropriate, excessive, or deficient. The ratio is a description of a relationship between two figures at a point in time, not a signal about direction or value.
Fee Retention: Take-Rate Divergence
Several assets in the dataset retain all or nearly all of the fees their networks collect. TRON, Hyperliquid, Canton, and Avalanche each show a take rate of 100% (annualized from the trailing 30 days), meaning revenue equals fees in the data as reported. Chainlink's take rate stands at 97.8% (annualized from the trailing 30 days). At the other end, Uniswap retains approximately 8.1% of its $1.26 billion in annualized fees (annualized from the trailing 30 days), passing the remainder to liquidity providers. Cardano retains roughly 20.0% of its fees, and BNB retains 10.0%. Where revenue is null—as it is for Bitcoin and Zcash—the dataset does not supply a split, so no take-rate comparison can be made for those assets. A high fee figure paired with a low take rate means the network or protocol intermediates a large volume of economic activity while capturing a smaller share of it as protocol-level income.
Chain-Basis vs. Application-Basis Figures Are Not Interchangeable
The accounting basis column in this dataset distinguishes between chain-level figures and application-level figures. Uniswap's fees are sourced from application:Uniswap V3; Chainlink's from application:CCIP. Every other asset draws from a chain-level source. Chain-level fees aggregate all activity settled on that ledger; application-level fees measure only what flows through a specific protocol deployed on top of some underlying chain. Comparing Uniswap's market-cap-to-fee ratio directly against Ethereum's, for example, conflates two different scopes of measurement. The application figure is a subset of broader on-chain activity and does not carry the same economic perimeter as a chain-level number. Readers should treat the two subgroups separately rather than ranking them in a single undifferentiated table.
Why These Are Not Equity Multiples
The single most important structural difference between a market-cap-to-fee ratio and an equity price-to-revenue or price-to-earnings multiple is that token holders generally do not hold a legal, enforceable claim on fee cash flows. In an equity context, revenue flows through a corporate structure to shareholders under defined legal rights. Token economics vary by protocol design, governance, and jurisdiction; the relationship between fee generation and any return to token holders is contingent, not contractual. Treating a market-cap-to-fee ratio as equivalent to a price-to-sales multiple therefore imports an assumption about cash-flow rights that the data does not support.
periods{'market_cap': 'point-in-time', 'fees_and_revenue': 'annualized from the trailing 30 days'}
observation date22 October 2026
Written by an AI language model from the figures listed above and nothing else. It contains no price view and no recommendation. Figures carry the reference periods shown on their own pages.
· claude-sonnet-4-6