How to Value a DeFi Token: A Guide to the Token Valuation Tool

A DeFi token is not equity. If a company earns a dollar, shareholders have a claim on it. A governance token has no such claim — revenue reaches a holder only if some mechanism explicitly routes it there, and several of the largest protocols in DeFi route nothing at all.

That is why the DeFi Token Valuation tool sorts every token into a capture class before showing a multiple, and why its headline figure is not P/E. This is a guide to reading it.

---

TL;DR — the five-minute workflow

1. Read the class badge before the multiple. Non-accruing means no mechanism routes revenue to the token; its P/E is a modelled hypothetical, shown with a dashed underline. Comparing it against a real one is a category error.
2. Read the net yield column, not the gross one. Gross is what holders are paid. Net subtracts the dilution they absorb. A wide gap between the two means dilution is the dominant fact about that token, and revenue growth will not outrun it soon.
3. Flip the revenue window from 7d to TTM. The gap between the readings is the signal. A 7-day figure far below the TTM one means the business is shrinking and the TTM number describes something that no longer exists.
4. Push the sell-through slider to 100%. If net yield survives the assumption that every unlocked token gets sold, the token is genuinely resilient. Most don't.
5. Find out what the price already assumes. In the screener, drag a token's next-cycle fees slider down until its bull case hits 0%. The line under the handle converts that to a percentage of today's fees — which is the fee recovery the current price already requires. Land near 100% of current and the market is pricing no recovery at all.
6. For a non-accruing token, switch on the fee-switch modeller. Set a take rate and payout, and you get the multiple that token would trade at if it ever routed revenue to holders. That is the actual bull thesis, priced.

At default assumptions, net holder yield is negative for a quarter of the set — including a token advertising a 14.4% gross yield.

---

Step 1 — the class decides whether the multiple means anything

A price-to-earnings ratio on a token with no earnings claim isn't a conservative valuation or an aggressive one. It's a category error: the number computes, and it means nothing.

Class	What it means	How to price it
Cash-flow	Revenue reaches holders today, with enough history to trust the policy	Like equity. The multiple carries a claim.
Transitional	Capture is live but young, capped, or revocable by a governance vote	Discount it. The mechanism is a decision, not a structure.
Non-accruing	No mechanism at all	Price is a reference. Any value case is a bet on a future fee switch.

Reading this table: rows are grouped cash-flow → transitional → non-accruing, then by market cap. The mechanism column names the actual route — fee share, buyback-and-burn, buyback-and-distribute, or none. A dashed underline on a P/E means the figure is modelled rather than measured. A small A / B / C letter beside a symbol is data confidence: A is clean, B carries one caveat (usually a manual override), C means no published unlock schedule or fee accounting that doesn't reconcile.

Two tokens can trade at the same multiple of the same revenue, and one is cheap while the other is a raffle ticket. That gap is the entire argument.

---

How every number here is built

• Income is annualized from a trailing window, never a single day: sum(window) × 365 ÷ window. One whale rebalancing moves a daily fee print by an order of magnitude. The default window is 90 days — long enough to survive a bad week, short enough to notice a business changing.
• Market cap is always circulating. Fully diluted valuation never appears as a valuation base anywhere in the tool: it prices tokens that don't exist yet against revenue that does, and double-counts dilution that gets handled explicitly later with better data. Circulating includes staked tokens the holder can exit; it excludes hard time-locks like ve-escrow, which genuinely cannot be sold.
• Income is split four ways, and only the last line is bankable:

Fees                       ← what users actually pay
 ├─ SupplySideRevenue      → LPs / depositors        (never reaches the token)
 └─ Revenue                → the protocol's take
     ├─ ProtocolRevenue    → treasury / team         (indirect at best)
     └─ HoldersRevenue     → buyback, burn, fee share  ← the only bankable line

Almost every "DeFi protocol revenue" league table quotes one of the top two lines. For a lot of protocols the bottom line is zero while the top one is enormous.

---

Step 2 — net holder yield

Three metrics get computed, and each is blind to something:

Metric	Formula	Blind to
P/E	market cap ÷ annualized holders revenue	The supply side. A cheap P/E on a heavily-emitting token is a mirage
Real yield	holders revenue ÷ market cap	The same. It's a payment rate, not a return
Net holder yield	(holders revenue − sellThrough × forward dilution) ÷ market cap	Nothing structural — this is the one to lead with

Gross yield answers how much is this protocol paying its holders? That's half a ledger. The other half is that the protocol is simultaneously printing tokens at you, and every new one dilutes the position you're being paid on. Netting them is arithmetic, not opinion.

Only one input is a judgement call, and it's a visible slider: sell-through, the share of next year's unlocks and emissions assumed to actually hit the market (default 60%).

Reading this table: sorted by net yield, best first. The bar is the net figure — right of centre is positive, left is negative. Aerodrome advertises 14.4% gross and lands at −0.3% net, because it mints roughly a quarter of its float every year through gauge emissions. Lighter pays 5.1% and nets −22.7%, with 46% of supply unlocking over the next twelve months. Mid-yield tokens with no emission overhang survive intact.

Forward dilution has two legs, and they add

forwardDilution = vesting12m + emission12m − burn12m

• Vesting — what a fixed allocation table releases: team and investor cliffs, airdrop tranches. Finite, and it ends. Lighter is almost pure vesting: 500M team and investor tokens on a cliff, no ongoing issuance.
• Emission — perpetual issuance outside that table: gauge emissions, block rewards, staking inflation. It does not end. Aerodrome is almost pure emission: no forward unlock table at all, but its Minter contract mints 21 basis points of total supply every week.
• Both — plenty of protocols do. Morpho vests investor tokens at 300,000 a day plus a separate grants programme.

These are researched from each protocol's own documentation, contract state, and governance records — not inferred from supply growth, which conflates redenominations and expiring locks with new issuance.

---

Step 3 — the upside decomposition

This looks like a price prediction and is not one. Start from an identity that is true by construction:

future price      future fees     future P/F      current supply
─────────────  =  ───────────  ×  ──────────  ×  ───────────────
current price     current fees    current P/F     future supply

                     torque      ×   re-rate    ×    1 / drag

No model, no regression. All the judgement in the exercise compresses into one term.

Factor	Measured as	Reads as
Fee torque	assumed next-cycle peak fees ÷ current annualized fees	How far below its own high-water mark the business sits
Multiple re-rate	P/F at its trailing-365d 90th percentile ÷ P/F today	How far below its own rich valuation the market sits
Supply drag	1 + (next-12m dilution ÷ circulating supply)	What dilution takes back from both

Two are measurements. Fee torque is your assumption, and it's the only slider.

Reading this table: every slider runs the same 0% to 500%, denominated in that token's own last-cycle peak. So two handles at the same position mean the same assumption, and the default sits at 80% of the peak everywhere — matching a previous high exactly is already the optimistic case.

The per-token part hasn't disappeared, it's on the line under each handle: 80% of the peak is 142% of current fees for HYPE and 78% for Morpho. Both readings are on screen at once, deliberately. Move a slider and the bull figure follows; a reset link appears once you've changed one.

The bear leg is not adjustable. It takes the 10th percentile of that protocol's own trailing two years — what it actually earned at its worst — and applies the same 20% haircut the bull default takes off the peak, because a bad quarter is evidence, not a floor. The point of a downside case is that it isn't your opinion.

Read the pair as brackets, not a range of likely outcomes. The bull case compounds two independent 90th-percentile events arriving together, so its joint likelihood is far below 10%.

---

The control panel

Everything on the left of the tool, and what moving it tells you:

Control	What it does	Why you'd move it
Revenue window	7d / 30d / 90d / 180d / TTM, each annualized	Compare two windows to see direction of travel
Cycle scenario	Rescales revenue to the 10th / 50th / 90th percentile of its own trailing 2 years	Ask "what does this look like at a cycle trough?" without inventing a number
Unlock sell-through	Share of next-year dilution assumed to hit the market (default 60%)	100% is the stress test; 0% assumes every unlocked token is held forever
Model a fee switch	Applies a hypothetical take rate and payout to tokens with no live capture	Price the bull thesis on a non-accruing token
Stress every payout	Forces that hypothetical payout onto all tokens, including those already paying	Compare every token on one payout policy
Next-cycle fees, vs last peak	Per-token slider in the screener, 0–500% of that token's own last peak, default 80%	Set your own fee assumption, or slide down to find the one the price already requires

Modelled figures never overwrite measured ones — they're marked with a dashed underline wherever they appear.

---

For how we assess the protocols underneath these tokens, see our protocol rating methodology. For the prior question of whether a yield protocol needs a token at all, see Does a Yield DeFi Protocol Even Need a Governance Token?.

---

This article is research and education, not financial advice. Every figure shown in the embedded modules is a trailing-window measurement of the past, and the upside decomposition is arithmetic applied to assumptions you set — none of it is a forecast. Do your own research before committing capital.