Token P/E Ratio
A token price-to-earnings ratio divides a token's valuation by the protocol's annualised earnings, meaning the revenue left once the token incentives and emissions the protocol spends to produce it are taken out. It tests whether the price is supported by cash flow rather than by emissions or narrative. Both inputs are choices: market cap or FDV on top, and fees, revenue or earnings underneath, and each choice moves the answer.
TradFi parallel: The equity P/E, with the same argument attached: the ratio only means something once everyone agrees which line of the income statement counts as earnings.
Key Takeaways
- 01Construction: valuation divided by annualised protocol income. On a 2023 snapshot, Lido's $1,700M FDV against roughly $35M of revenue (5% of staking rewards, after node operators took half of a 10% fee) gave a multiple near 48.6
- 02Fees, revenue and earnings are three different denominators: fees are what users pay, revenue is what is left after the supply side takes its cut, and earnings are revenue after token incentives and emissions
- 03The choice can move the answer by an order of magnitude: UNI screened at 88.8 times sales but 2.8 times fees on the same July 2026 snapshot
- 04The numerator encodes supply: HYPE cost 15.5 times on market cap and 70 times on FDV, because only about a fifth of its supply was circulating
- 05The ratio moves on assumptions as much as on price: Lido's 48.6 fell to 37.8 purely from an assumed growth path in staked ETH, with prices held constant
- 06Earnings only matter to a holder if a mechanism routes them to the token: LDO had no buyback and no burn as of the 2023 analysis, and Ethena's sENA fee switch had its conditions declared met in September 2025 with activation still unconfirmed
How It Works
The construction is standard, but crypto forces you to build the denominator yourself. Tokenomist's Lido analysis, run on a mid-to-late-2023 snapshot, is a clean worked example. Lido charged a 10% fee on staking rewards and paid half of it to node operators, so the protocol kept 5% of staking rewards. With roughly 8.15 million staked ETH generating roughly 399,000 ETH of staking rewards at an assumed 4.90% staking APR, Lido's 10% fee came to about 40,000 ETH, of which half went to node operators and the protocol kept around 20,000 ETH, about $35 million at a $1,750 ETH price. Set against a $1,700 million fully diluted value, the resulting multiple was approximately 48.6. Node operators are supply side, so in DefiLlama's terms that is a revenue multiple, which is the closest available proxy for a P/E.
The denominator is where the disagreement lives. In crypto, fees are the total users pay and revenue is only the slice the protocol keeps, with the rest going to liquidity providers, lenders, stakers or other suppliers of capital. Earnings sit a step below revenue again, netting out the token incentives and emissions the protocol spends to generate it. Price-to-sales divides market cap by trailing twelve-month protocol revenue and is the closest thing to a P/E; price-to-fees uses total fees instead and is a steadier basis where the revenue split is murky. The gap between them is not cosmetic. On a July 2026 snapshot, UNI screened at 88.8 times revenue, on an annualised run-rate basis, and 2.8 times fees, because almost all of what users pay flows to liquidity providers rather than to the protocol. Quoting a token P/E without naming the denominator says almost nothing.
The numerator decides whether the ratio prices today's float or all future supply. HYPE cost 15.5 times on market cap but 70 times on FDV in the same snapshot, because only about a fifth of its supply was circulating. That spread is the unlock overhang the headline multiple hides, which is why FDV-to-revenue is worth carrying alongside price-to-sales for any token with a long vesting tail. The Lido case shows the other sensitivity: holding price and APR constant and assuming Lido kept a 35% share as staked ETH grew, the multiple fell from 48.6 to 37.8 on an assumption alone, with nothing about the token changing.
A low multiple is not the same as a cheap token. Across the buyback-and-burn cohort Tokenomist examined, almost every program was running on falling revenue, which makes today's multiple more expensive tomorrow unless the cash flow returns. Two structural caveats matter more than the arithmetic. First, earnings only reach holders if something routes them there: at the time of that 2023 analysis LDO was a governance token with no buyback and no burn, so protocol income never touched the token directly, and Lido only adopted a capped buyback framework, a repurchase rather than a burn, in November 2025. Ethena kept a share of USDe yields as revenue while the fee switch to sENA holders waited on activation conditions, which Ethena declared met in September 2025 without activation being confirmable as of August 2026. Second, emissions sit on the other side of the ledger. Tokenomist's own analysis read Ethena's emission-to-revenue ratio as running high, which is the sanity check a P/E on its own will never give you.
Real World Examples
Lido: A Multiple Built on What the Protocol Kept
View →On a mid-to-late-2023 snapshot, Lido charged 10% on staking rewards and paid half to node operators, keeping 5%. Around 8.15 million staked ETH at an assumed 4.90% APR produced roughly 399,000 ETH of rewards, so the 10% fee came to about 40,000 ETH and Lido's own share to around 20,000 ETH, about $35 million, against a $1,700 million FDV for a multiple near 48.6. Projecting Lido holding a 35% share as staked ETH grew brought it to 37.8 without any price change.
Uniswap: The Same Token at 88.8x and 2.8x
View →On a snapshot dated 2026-07-29, UNI carried a $2.42B market cap against $27M of annualised protocol revenue, a run-rate figure rather than a trailing twelve months, since Uniswap only switched its protocol fee on in December 2025. That gave 88.8 times sales but only 2.8 times fees. The gap is structural: most of what Uniswap users pay goes to liquidity providers, so the fee base and the revenue base describe two very different businesses.
Hyperliquid: 15.5x on Market Cap, 70x on FDV
View →HYPE showed a $12.2B market cap against $784M of trailing revenue in the same snapshot, for 15.5 times sales. Measured on fully diluted value the multiple rose to 70 times revenue, because only about a fifth of the supply was circulating. The spread between the two numbers is the unlock overhang the headline multiple leaves out.
Ethena: Revenue Without a Payout Yet
View →Ethena kept 20% of USDe yields from perp funding, ETH staking and T-bill returns, and that retained share is the revenue a P/E would divide into. The approved fee switch would route part of it to sENA holders, and Ethena declared the activation conditions met in September 2025, though activation could not be confirmed as of August 2026. Tokenomist's analysis also read the protocol's emission-to-revenue ratio as running high.
Aster: No Denominator to Divide By
View →ASTER appeared in the same valuation table with a $1.67B market cap and no price-to-sales figure at all, because DefiLlama did not track its revenue. Only price-to-fees could be shown, at 3.8. A missing denominator is common enough that any screen built on token P/E needs an explicit rule for what to do with the gaps.
Frequently Asked Questions
What is the difference between a token P/E and price-to-sales?
Price-to-sales divides market cap by trailing twelve-month protocol revenue and is the closest crypto equivalent of a P/E. Revenue is already net of the supply side, the liquidity providers, lenders, stakers or node operators who generate it, which is what Tokenomist's Lido analysis captured by counting only the 5% of staking rewards Lido kept after node operators took their half. A true P/E goes one step further again and nets out token incentives and emissions. Price-to-fees is a third variant that uses total user-paid fees, useful where the revenue split between the protocol and its suppliers is unclear.
Should I use market cap or FDV in the numerator?
Use both. Market cap prices the float you can trade today; FDV prices the supply that will exist once vesting completes. HYPE illustrates the gap, at 15.5 times revenue on market cap and 70 times on FDV with roughly a fifth of supply circulating. For a token deep into its unlock schedule the two converge, and for a low-float launch the FDV version is the one that carries the dilution.
Does a low token P/E mean the token is cheap?
No. Cheap and good are not the same thing. In the buyback-and-burn cohort Tokenomist tested, almost every program was running on falling revenue, so a low trailing multiple was often a shrinking numerator chasing a shrinking denominator. Check the revenue trend, the supply still to unlock, and whether any mechanism actually routes earnings to holders before treating a low multiple as a discount.
Why do some tokens have no P/E at all?
Two reasons. Some tokens have no claim on protocol income: as of the 2023 analysis LDO was a governance token with no buyback and no burn, so earnings existed at the protocol level but never reached the token directly. Others simply lack data, as with ASTER, whose revenue was not tracked by DefiLlama, leaving only a price-to-fees figure. In both cases the absence is information, not an error to be filled in with an estimate.
Related Terms
Track on Tokenomist
Supply-side analysis for educational purposes. Not financial advice. Verify assumption and precision labels on the relevant token page.
Tokenomist.ai provides a complete solution for supply-side tokenomics data. Analyze future token emissions, track vesting schedules, and compare standardized tokenomics and allocation across projects to gain actionable insights