Market-cap-to-TVL ratio

What the ratio is trying to ask

The intuition borrows from equities. If a protocol holds a large pool of user deposits and its token is priced cheaply against that pool, perhaps the market has not yet priced the activity those deposits generate. If the token is priced far above the deposits, the market is paying for growth that has not arrived.

The analogy is loose in a way worth naming. Deposits in a DeFi protocol are not the protocol's assets. They belong to users, who can withdraw them. Token Terminal is careful about this in its own definition of TVL, describing it as user deposits to the protocol's smart contracts and noting explicitly that these assets are not owned by the protocol.¹ So the denominator is not a balance sheet. It is other people's money passing through.

That does not make the ratio useless. It makes it a comparison between a valuation and a volume of activity, which is a reasonable thing to look at as long as nobody calls it a book value.

Both inputs are contested, so the ratio inherits both disputes

CoinMarketCap publishes the ratio directly on its coin pages. For Ethena on 3 August 2026 it showed a market cap of $880.15 million, a TVL of $4.09 billion, and a market cap to TVL ratio of 0.2149.² Compute the same ratio from CoinGecko's API read the same day, with a market cap of $880,036,871 and a TVL of $3,855,382,456, and you get 0.228.³

The market caps were within about a hundredth of a percent of each other. Essentially the entire difference in the ratio comes from the TVL denominator, where the two providers were roughly $235 million apart on the same protocol on the same day.

Six percent is small enough to ignore in isolation and large enough to matter when you are ranking a set of protocols by this ratio and the gaps between them are of that order. If you build a comparison table, pull every figure from one provider on one day and say so in a footnote.

Not every provider publishes it, and the absence is informative

Token Terminal, which tracks TVL and market capitalization for the same protocols, does not publish a named market cap to TVL ratio at all. Its valuation section offers price to fees and price to sales ratios, on both fully diluted and circulating bases, computed against annualised fees and annualised revenue.¹

That is a deliberate choice and a defensible one. Fees and revenue are things the protocol earns. Deposits are things users park. A multiple built on earnings answers what the token is priced at relative to what the protocol produces; a multiple built on deposits answers what the token is priced at relative to how much capital happens to be sitting there.

Our position matches Token Terminal's here. If you have fee data, use a fee multiple. Use market cap to TVL when fee data is unavailable or unreliable, and treat it as the weaker instrument it is.

The academic result that should temper any enthusiasm

A 2025 paper by Matt Brigida tested whether TVL carries information about returns at all. Portfolios sorted on TVL were found to produce returns explained by aggregate crypto market performance, meaning they can be replicated with weighted positions in the crypto market portfolio and can be priced with standard asset pricing tools. The result held for total TVL and for a TVL measure adjusted to remove staking, pool2, governance tokens, borrows, double counting, liquid staking and vesting.⁴

The paper is careful about what this does and does not mean. It notes that TVL may still proxy for protocol usage or perceived utility. It just does not appear to carry information that standard factors do not already contain.

Read that as a ceiling on the ratio's ambition rather than a reason to discard it. As an operating diagnostic for one protocol tracked over time, market cap to TVL still tells you something about how the market is repricing your deposits. As a screening tool for picking between protocols, the evidence is not there.

Comparing across verticals produces nonsense

A lending market's deposits are its product; capital sitting idle earns nothing for anyone, so TVL and revenue move roughly together. A decentralised exchange can generate large fee income on a comparatively small pool if turnover is high, so its ratio will look expensive against a lender's on identical economics. A liquid staking protocol accumulates enormous TVL by construction and takes a thin cut of it. A governance token for a protocol that holds no user deposits at all has a denominator near zero and a meaningless ratio.

So the comparison is only legible within a vertical, between protocols with similar fee structures, at the same point in time, from the same provider. That is four constraints, and most published comparisons satisfy none of them.

The variant we find more useful in practice is fee yield on deposits: annualised protocol fees divided by average TVL. It answers the question people are actually asking, which is whether the deposits do any work.

How we use it in an audit

We compute it, we look at it, and we never conclude anything from it on its own. What it does well is prompt the right follow-up. A very low ratio raises the question of why the market is discounting a protocol holding that much capital, and the answer is usually visible in fee capture, token holder rights or a governance risk somebody has already noticed. A very high ratio raises the question of what the token is priced against, and the answer is usually an expectation about a product that has not shipped.

Neither reading is a valuation and neither is a recommendation about any asset. The ratio narrows where to look. The work still happens in the fee statements, the contract permissions and the business underneath, because a token is infrastructure for a business and no ratio built from deposits can tell you whether that business works.