What the fee-generating assets look like on 19 September 2026

The spread of market-cap-to-fees ratios

As of 19 September 2026, the market-cap-to-fees ratios across the fourteen assets in this dataset span from 2.4× (Uniswap, annualized from the trailing 30 days) to 23,860× (Zcash, annualized from the trailing 30 days)—a range of roughly four orders of magnitude. A wide ratio indicates that each dollar of annualized fee activity is associated with a large amount of market capitalization; a narrow ratio indicates the opposite. Crucially, neither end of that spectrum carries an inherent normative interpretation: the ratio describes a relationship between two figures, not a direction of travel or a mispricing signal.

Fee retention: pass-through versus capture

The dataset distinguishes fees from revenue where data permit. TRON and Hyperliquid each report a take rate of 100% (annualized from the trailing 30 days), meaning all fees recorded on those chains are counted as protocol revenue. By contrast, Uniswap reports a take rate of approximately 8.2% (annualized from the trailing 30 days)—the remainder flows to liquidity providers rather than the protocol itself. Ethereum sits at roughly 23% (annualized from the trailing 30 days). Assets with high take rates retain more of gross fee flow in the protocol layer, while low-take-rate protocols redistribute most economic activity to other participants. Neither structure is presented here as superior; they reflect different design choices about value distribution.

Chain-basis versus application-basis figures are not comparable

Several entries—Uniswap, Chainlink, and Rain—are measured on an application basis, meaning fees are those attributable to a specific smart-contract application. The remaining entries are measured on a chain basis, capturing aggregate activity across an entire network. An application running on Ethereum, for instance, contributes to Ethereum's chain-level fee total and reports its own application-level fee total separately. Summing or directly ranking chain-basis and application-basis figures against one another would double-count activity and produce misleading comparisons. Readers should treat the two subsets as parallel but distinct series.

Why these multiples are not equity multiples

The single most important caveat is that token holders do not automatically possess a legal or contractual claim on fee cash flows in the way equity shareholders possess a claim on corporate earnings. A market-cap-to-fees ratio may superficially resemble a price-to-sales multiple, but the analogy breaks down because the economic rights encoded in a token—governance participation, fee discounts, burn mechanisms, or none of the above—vary by protocol and are defined by on-chain rules rather than corporate law. Furthermore, Bitcoin reports no revenue figure at all (the JSON field is null), illustrating that fee generation and any form of distributable income are not equivalent concepts across this asset class. The ratios here describe a market-cap-to-activity relationship; they do not measure profitability, distributable cash flow, or any quantity that maps cleanly onto traditional valuation frameworks.