Market cap to annualized fees, worked through step by step
The arithmetic takes one line; the choices hidden inside it are where two careful analysts end up with different numbers for the same network.
Market cap to annualized fees divides what the market values a network's units at by the fees users paid to use it over a year. It is the closest thing the sector has to a price-to-sales ratio, and the resemblance is superficial enough to be dangerous. Fees are not sales, the denominator is not audited, and the result is not a price for anything a holder owns.
The formula is market cap ÷ annualized fees. Everything interesting is in how each side is built.
Building the numerator
Market capitalization is price multiplied by a supply figure, and the supply figure is a choice. Circulating supply attempts to count units that are transferable today, excluding locked, unvested and unminted tokens. Total supply counts everything that exists, including locked allocations. Maximum supply counts everything that will ever exist, where such a cap is defined at all.
Different data providers apply different tests for what counts as circulating, and the treatment of foundation holdings, unclaimed airdrops and long-dated vesting contracts varies between them. This site states its rule on the methodology pages and shows circulating supply alongside total supply so the gap is visible rather than assumed away. Where the gap is large, every ratio in this track has a second version built on the fully diluted figure, covered later in the track.
Building the denominator
Protocol fees are the total amount users paid to transact, borrow, swap or otherwise use the system, before any split between the parties who receive it. On a base-layer network that is the sum of transaction fees. On a decentralized exchange it is the swap fee charged on volume. On a lending protocol it is interest paid by borrowers. These are different economic activities being added into one line, which is the first thing to hold in mind when the resulting number is compared across categories.
Annualizing then requires picking a window. Each choice trades noise against staleness, and the trade is not neutral.
| Window | Construction | What it captures | Main distortion |
|---|---|---|---|
| 24 hours | fees_24h × 365 | Today | A single busy day can multiply the annual figure several times over |
| 7 days | fees_7d × 52.14 | Current week, weekday effects smoothed | Still dominated by one event if that event lasted days |
| 30 days | fees_30d × 12.17 | Current regime | Extrapolates a market condition that may not persist |
| Trailing 365 days | Sum of the last year | A full range of conditions | Reacts slowly; a network that changed in month two still looks like its old self |
This site publishes 24-hour, 7-day and 30-day fees separately, and an annualized fees figure whose window is stated rather than implied. Comparing an annualized number built from one window against one built from another is a common and entirely silent error, because both are labeled the same way.
The arithmetic, with illustrative figures
The numbers below are invented for the demonstration and correspond to no asset. Suppose a network has 400 million circulating units at a price of 5 dollars, giving a market cap of 2 billion dollars. Suppose users paid 1.2 million dollars in fees over the last 30 days.
- Annualize the fees:
1,200,000 × 12.17 = 14,600,000dollars. - Divide:
2,000,000,000 ÷ 14,600,000 ≈ 137. - State it precisely: the network's circulating units are valued at roughly 137 times the fees paid to use it during the last 30 days, projected forward a year.
Now change one input. If the same fee total had been earned in a month containing an unusual event, and the following month returned to 600,000 dollars, the identical calculation yields roughly 274. Nothing about the network changed between those two readings. The fee growth series exists to make that kind of swing visible before the ratio is interpreted.
A second sensitivity runs through the numerator. If the same network had 900 million units outstanding of which 400 million circulate, the fully diluted version of the same calculation produces roughly 308 at an unchanged price. The network is identical; the ratio has more than doubled because a different supply definition was used.
What the number does not mean
A price-to-sales ratio for a company implies that sales eventually become cash available to shareholders, after cost and risk. Nothing of the sort holds here. Most of the fees in the denominator are paid to somebody who is not a holder: to validators, miners, or liquidity providers, all of whom are supplying a resource rather than owning the network. The portion that reaches holders, if any, is a separate measurement, published as market cap to revenue and covered in the next article.
The ratio also has no natural resting level. Equity multiples are anchored by decades of observed outcomes, by the cost of capital, and by the fact that a company which stops generating cash eventually stops existing. A network with almost no fees can persist indefinitely, because block production does not depend on fee income in the way a payroll depends on revenue. There is no arithmetic level at which the ratio becomes correct.
Practical reading
The ratio is most informative as a time series for a single asset, where the supply definition and the fee boundary are held constant and only the market's assessment and the network's usage vary. It is least informative as a single cross-sectional snapshot ranking unlike networks against each other, which is the form in which it most often appears.
Two supporting checks are worth running alongside it. The first is whether fee revenue is concentrated in a handful of days, visible in the gap between the 24-hour and 30-day series. The second is whether the fees are being paid by genuine users or generated by activity that exists to produce the appearance of usage, a pattern discussed under wash trading and more common on venues with weak surveillance than on public ledgers.
The fees pages carry the underlying series and the window used for each asset. The next article takes the same structure and narrows the denominator from all fees to the portion that arguably accrues to the protocol.