Every project on this site carries one of three verdicts about whether it is worth holding for the next ten years.
Hold. A project likely to still be here in a decade, with a token that does real work and no finding that its holders are being extracted from, misled, or replaced by insiders.
Watch. A project that could go either way over the same horizon. The business may be real, but how the project is owned, run, or growing sits in a gray area that keeps the case from settling.
Avoid. A project where specific reasons not to hold have been found. Either the token has no clear reason to exist, or the way it works or is governed carries a defect that survives every argument in its favour.
None of these verdicts is a price call: a Hold does not mean buy now, and an Avoid does not mean sell now. Each is a judgment about the token itself, not the company or the technology behind it. A company can sign a client every week, and none of that revenue has to reach its token; a technology can be sound while the token routed through it has no clear job.
The token is what a person is buying, and the token is what these pages judge.
Why ADA exists
Every transaction on Cardano costs a fee, paid in ADA, which stops the network from being flooded with free transactions that would slow it down for everyone.
ADA is also the collateral that secures the network. Cardano uses proof of stake, meaning the computers that confirm transactions must lock up ADA to be allowed to do that work, and that locked stake is what they stand to lose if they try to cheat. Anyone holding ADA can delegate it to one of those operators and share in the rewards, without giving up control of the tokens themselves.
Holders also vote on changes to how the network runs.
Each of these is a real use that has nothing to do with anyone expecting the price to rise: the token is consumed by using the network, it is the collateral securing it, and it carries a vote.
The money is real, and it is small
Over the trailing year, people paid $1,284,119 in fees to use Cardano. That figure is measured cleanly, covering 99.7% of the days in the year with complete daily data going back to September 2018.
Against Cardano's size, that number is strikingly small. Divide the year's fees by the total value of all ADA in existence and the network's cut of its own economy comes to 0.0150%, about one part in seven thousand. For comparison, an app store typically takes fifteen to thirty percent of what happens on it. Even by the standards of other blockchains, which charge far less than that, Cardano is at the very bottom of the range.
That is worth reading in the direction it points: on the question of whether Cardano extracts value from the people using it, the measured answer is no, emphatically. Nothing in the data suggests a network taking more than its share. It charges almost nothing.
Source: Cardano fees, DefiLlama
What Cardano takes from its own economy
Yearly fee income as a share of the network's total value, against a typical app store's cut. Log scale, because the two differ by roughly three orders of magnitude.
Eight years of fees, and a long decline
Cardano's fee record runs back to September 2018, one of the longer continuous records in this assessment, and across that whole history the network has taken in $27,630,328.
The shape is a boom and a long fall. Fees climbed through 2021 to a peak of $1,716,691 in a single month in November of that year, during a period of intense speculative activity across all of crypto, and a single day that month brought in $161,530. Activity fell back through 2022 and 2023, recovered partially in 2024, and has been declining steadily since.
The decline has not stopped. Measured over the last three months and projected across a year, the pace comes to $500,907, against a trailing-year total of $1,284,119, a fall of about two and a half times. Recent months have run between $38,000 and $63,000, and the most recent day on record took in $1,031, which is 99.4% below that 2021 peak day.
Source: Cardano fees, DefiLlama
Cardano's monthly fees, 2018 to 2026
Total transaction fees paid each month across the network's full eight-year record. A 2021 peak, a partial recovery through 2024, then a steady decline to under $40,000 a month.
Where the money goes, and where it stops being traceable
Of the fees Cardano collects, the overwhelming majority goes to the operators who run the network and the people who delegate their ADA to them. Over the trailing year that share came to $1,166,419, about 91% of all fees paid.
The remaining 8.4%, $107,821, goes somewhere this assessment cannot follow. The data that would identify the recipient does not exist for Cardano: the two channels that would report it return errors rather than figures, and no public record characterising the token's rights fills the gap.
That means something specific, and it is worth being precise. It is not that the money vanishes, and it is not that Cardano is hiding anything. It is that the measurement simply is not available. Cardano is widely understood to route these fees to a reward pot and treasury rather than destroying them, but that description is not evidenced by anything in the material this assessment is built on, so this page does not assert it as fact.
What can be said: the 91% is measured and reaches the people securing the network. The remaining 8.4% is measured as an amount and unknown as a destination.
Source: Cardano fees, DefiLlama
The price, against the earning
Cardano's market value, the price of every ADA in existence multiplied together, is $8,564,124,543. Divide that by the annualised fee run-rate from the most recent three months and Cardano trades at roughly 17,097 times its current yearly fee income.
For comparison, a wireless network project this framework has used as its example of an extreme, speculative valuation traded at roughly 87,000 times its own fees. Cardano is five times better than that, but it is one of the widest gaps in this assessment, and at seventeen thousand times current earnings the price is built on what Cardano might become rather than what it currently takes in.
Two things make that number harsher than it has to be, and both should be stated. The denominator deliberately uses the recent, declining three-month pace rather than the trailing year; on the trailing-year figure the multiple would read closer to 6,669 times. And it prices the chain by itself, excluding every application built on top of it. There are 53 genuine on-chain projects running on Cardano and none of their activity is counted here.
Cardano's price against what it earns in fees
Market value divided by yearly fee income. Cardano trades at roughly 17,097 times its annualised fees. This prices the chain alone, excluding the 53 applications built on top of it.
The applications built on Cardano have collapsed
Beyond the network itself, a set of applications runs on top of Cardano holding user funds: lending platforms, exchanges, and other financial tools. The total value held inside them is a standard measure of how much real activity a network's ecosystem is getting.
That total peaked at $721,034,286 in December 2024. Today it stands at about $61.2 million, a fall of 91.5% from the peak, measured across 1,703 daily observations rather than two isolated snapshots.
This and the fee decline described earlier are two independent measurements of different things. One counts capital parked on the chain; the other counts transactions people actually paid for. They are not the same quantity, they are drawn from different sources, and they point the same direction. When two unrelated measures agree this closely, the picture they describe is harder to dismiss as an artifact of either one.
Source: Cardano TVL, DefiLlama
Value held in applications built on Cardano
Total funds held inside Cardano's applications, January 2022 to August 2026. A three-year build to a December 2024 peak, then a sustained fall of more than ninety percent.
The November 2025 network split
In November 2025, a single malformed transaction triggered a bug that had been sitting dormant in Cardano's software since 2022, and the network split in two: one chain processing the bad transaction, one rejecting it.
Recovery took about fourteen and a half hours. There was no central switch to flip and no rollback of the ledger. The operators running the network upgraded their software voluntarily, one by one, until the network converged back onto a single chain. Intersect, the organisation maintaining the core software, published a detailed account afterward stating plainly that no funds were stolen and that the healthy chain had maintained a consistent record throughout.
This reads in two directions and both are worth holding. A dormant bug surviving three years in software that secures billions of dollars is a real failure, and the split it caused was real. The recovery is also a real demonstration of something Cardano claims about itself: there was no authority who could unilaterally fix it, so the fix required the operators to agree, and they did, in under a day.
Source: Intersect MBO, 23 November 2025
What Hold means for someone holding ADA
Cardano charges almost nothing for what it does, and the two measures of how much it is used have both collapsed by more than ninety percent.
The case in its favour is measured rather than asserted. The network's cut of its own economy is 0.0150%, orders of magnitude below anything resembling extraction, and about 91% of what it does collect goes to the people securing the network. Across an entire lens of tests built specifically to catch a project's stated claims diverging from its actual behaviour, Cardano returns no failure at all: nothing in the measured data suggests it takes more than it says it does. Its fee history is eight years long and complete, its code is open and actively maintained, and when its network split in 2025 the recovery happened by operator consensus rather than by anyone's authority.
The case against it is smaller in count but sharper. Two tests fail, and both are judgments about what Cardano is rather than how it performs: that it came out of a funded research organisation rather than emerging from the margins, and that it launched polished and serious rather than as something easy to dismiss and underestimate. Neither rests on a number, and a reader who reads academic rigour as a strength rather than a disqualifier is disagreeing with the entire case against. Separately, and this is not a judgment call: the value held in applications on Cardano is down 91.5%, and its fee income is down two and a half times and still falling, while the price sits at seventeen thousand times what the network currently earns.
Hold does not mean buy now. It means the case for holding ADA over the next ten years rests on economics that are clean but very small, against usage that has fallen a long way and has not turned around.
Someone who owns ADA owns a claim on a network that takes almost nothing from its users and passes most of what it takes to the people securing it, whose academic origins this assessment marks against it and a reader might not, and whose usage has fallen further than almost anything else measured here.
ROI through the years
Bought and held to 9 September 2026, never traded. CARDANO is younger than the 10-year window, so that row is everything since it started trading instead.
Bought at listing, April 2018
−12%
Bought 5 years ago, September 2021
−92%
Bought 3 years ago, September 2023
−16%
Returns and falls are measured on daily closing prices, so a one-minute crash that did not settle cannot enter them. The worst drop is the largest peak-to-trough fall inside that row's window, not the token's all-time worst. Bars sit on a log scale of what one dollar became. Sources: Binance.
What could not be answered
Who receives the 8.4% of fees that does not go to network operators. The channels that would report this return errors rather than figures, and no public record of the token's rights fills the gap. The amount is measured; the destination is not.
How much new ADA is created each year to pay the people securing the network. On a network like Cardano this is typically the largest single flow in its economics, larger than fees, and no measurement of it was available for this assessment. Every figure on this page excludes it.
How many transactions Cardano processes, and how fast. The address-level data these measures normally come from was never collected for Cardano, so its usage is assessed here on fee income and value-held alone.
Whether Cardano has been independently audited. No audit could be confirmed one way or the other, and unusually, the word "audit" does not appear anywhere in Cardano's own documentation either. This is an absence of information, not a finding that no audit exists.
Anything Cardano itself has published about these questions. Cardano's own disclosures were never gathered for this assessment. Where this page says something is unknown, that means unknown to this assessment, not that Cardano has declined to say.
Karinva takes no payment from any project it covers.