What the fee-generating assets look like on 23 October 2026
Fee-stream data as of 23 October 2026 reveals wide dispersion in market-cap-to-fee ratios and retention rates across fourteen digital assets.
Spread of Market-Cap-to-Fee Ratios
Across the fourteen assets observed on 23 October 2026, the market-cap-to-fees multiple (fees annualized from the trailing 30 days; market cap point-in-time) spans more than four orders of magnitude. At one end, Uniswap records a ratio of 3.69× against annualized fees of $1.22 billion; at the other, Bitcoin registers 18,832× against annualized fees of $87.3 million, and Zcash reaches 17,377× against just $1.2 million. A wide ratio indicates that a large market cap coexists with a small annualized fee base as measured over the trailing 30 days — it does not, by itself, indicate anything about whether that relationship is warranted, since fee generation is only one of many factors that market participants may weigh.
Fee Retention: Pass-Through vs. Capture
The data distinguish between total fees paid by users and revenue retained by the protocol or validators. TRON and Hyperliquid each report a take rate of 100%, meaning all fees recorded on-chain are classified as revenue. Canton and Avalanche show the same pattern. By contrast, Uniswap retains only 8.1% of its $1.22 billion annualized fee base as revenue ($99.6 million annualized from the trailing 30 days), with the remainder flowing to liquidity providers rather than the protocol treasury. Ethereum sits between these poles at a 36.0% take rate, while Cardano retains just 20.0% of its already modest fee base. Comparing gross-fee multiples across assets with very different retention structures can be misleading; a low multiple on gross fees may correspond to a much higher multiple on retained revenue, or vice versa.
Chain-Basis vs. Application-Basis Figures
Several entries in this dataset are accounted for on a chain basis (e.g., chain:Ethereum, chain:Solana), while two — Uniswap (application:Uniswap V3) and Chainlink (application:CCIP) — are measured at the application layer. Chain-basis figures aggregate all fees paid to secure or use the underlying network; application-basis figures capture only activity attributable to a specific protocol running on top of one or more chains. These are structurally different units of account: an application's fees are a subset of the chain traffic it generates, while a chain's fees include activity from all applications deployed on it. Direct ratio comparisons between Chainlink's 139× multiple and Ethereum's 1,389× multiple, for example, must be interpreted with this accounting difference in mind.
Why These Are Not Equity Multiples
The most consequential reason these market-cap-to-fee ratios cannot be read as equity price-to-earnings equivalents is that token holders generally have no enforceable residual claim on fee cash flows. Unlike equity, where a share represents a legal ownership interest with defined distribution rights, holding a network token typically confers governance influence, staking eligibility, or utility access — none of which automatically entitles the holder to a proportional share of fees. Fee revenue may be burned, redistributed to validators, routed to a treasury, or distributed in ways that vary by protocol and may change through governance. The figures here describe economic activity on these networks as annualized from the trailing 30 days to 23 October 2026; they do not describe cash flows accruing to token holders in the way earnings accrue to equity shareholders.
periods{'market_cap': 'point-in-time', 'fees_and_revenue': 'annualized from the trailing 30 days'}
observation date23 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