TVL, APY, and FDV: How to Read DeFi Metrics Like an Analyst

TL;DR

• TVL, APY, volume, and FDV each measure something real and each gets dressed up before it reaches you. Knowing the honest signal and the standard lies of each one is the core analyst skill.
• TVL honestly signals trust at scale, but it double-counts dollars through protocol stacks, swells when token prices rise, and can be rented with emissions.
• APY is the taller spelling of APR, and a yield number with no time direction attached (trailing or projected) is an ad, not a rate.
• Volume is the most gameable number on the board because trading with yourself is cheap. Fees actually paid are the closest thing to a receipt.
• The method fits in three questions: what does it measure, who benefits from it being big, and what does it cost to fake?

How do you read DeFi metrics like an analyst?

Reading DeFi metrics like an analyst means knowing, for each headline number, three things: what it actually measures, who benefits when it looks big, and how expensive it is to fake. Numbers that are costly to fake, like fees users actually paid, outrank numbers that are nearly free to fake, like trading volume. That is the entire method. The rest of this article applies it to the four numbers on every protocol's front page: TVL, APY, volume, and FDV.

Here is why you need it. Someone will pitch you a protocol this year: a friend, a podcast, a stranger with conviction. The pitch will be numbers. A billion locked. Forty percent yield. Huge volume, tiny market cap. Watch an insider hear the same pitch and notice what happens before the sentence ends: a tab is open. Almost always it is DefiLlama, the dashboard the whole industry checks first. This article is that tab, decoded.

One image will carry us the whole way: a restaurant with a line out the door. From the street, the line is all you can see, and it looks like proof. But a line is knowable. You can ask who is standing in it, whether they were paid to stand there, and whether the same people quietly rejoin it every hour. Every DeFi metric is some version of that line, and every question you can ask about the line has a metric-shaped answer.

What does TVL actually measure?

TVL, total value locked, is the market value of everything parked in a protocol's smart contracts. It is the biggest number on the board and the first one every pitch reaches for.

Its honest signal is worth respecting: people trusted this thing with real money, at scale. Capital is skeptical, and a billion dollars sitting in a contract means a lot of skeptical people were persuaded. TVL is the length of the line outside the restaurant.

Now the three ways it lies.

Lie one: it double-counts dollars through stacks. DeFi protocols stack: you deposit ETH into a staking protocol, take the receipt token to a lending market, borrow against it, and deposit the borrowed asset somewhere else. That is one original dollar, and every floor of the tower counts it in its own TVL. Summed across protocols, the industry's total TVL is meaningfully larger than the actual money in the room. One diner can stand in three lines at once. (How these stacks work, and what each layer risks, is its own article: money legos and DeFi composability.)

Lie two: it swells when token prices rise. TVL is measured in dollars, so if a protocol holds mostly ETH and ETH doubles, TVL doubles with zero new users and zero new deposits. The line got richer, not longer. A TVL chart during a bull market mostly shows you the market, not the protocol.

Lie three: it can be rented. Protocols print their own tokens to pay people for depositing, a practice covered in depth in where DeFi yield comes from. Emissions-rented TVL is a paid crowd: real money, but it arrived with the incentives and it leaves with them.

That third lie suggests the analyst's TVL question. Two protocols, identical 500 million TVL. One is mostly stablecoins that arrived this month, farming emissions. The other is unincentivized ETH that has sat for a year. The dashboard prints the same number for both, but they are opposite stories, because money that stays without being paid is the strongest signal money can send. A line of paid extras and a line of paying regulars fill the same street.

What is the difference between APR and APY?

The yield number is the one beginners trust most and inspect least, so learn its two spellings.

APR is the simple rate: the stream of rewards, stated plainly, per year. APY is the same stream assuming every payout is instantly reinvested, compounding on itself. Same farm, same dish, two menus: a 20 percent APR spells itself as roughly 22.1 percent APY. Nothing dishonest has happened yet, both numbers are arithmetically true. But marketing quotes the taller spelling, every time, and the gap widens as rates climb. When you compare two yields, first make sure they are spelled the same way.

The second dressing is the one that actually costs people money: trailing versus projected. Trailing yield is yesterday's weather: what the pool actually paid over the last day, week, or month. It may not repeat, but at least it was measured. Projected yield is a forecast written in the seller's handwriting: what the pool would pay if current conditions held, or if optimistic assumptions came true.

The rule: a yield number with no time direction attached is an ad, not a rate. Before you trust any APY, ask two questions. Is this trailing or projected? And, from the yield-sources playbook, what is actually paying it? A 38 percent APY paid mostly in the protocol's own freshly printed token is a very different object from 38 percent paid in fees, even though the dashboard renders them identically.

Why is trading volume the most gameable metric?

Volume, how much traded in a day, should measure activity. In practice it is the cheapest number on the board to fake, because on venues with low fees, trading with yourself costs almost nothing. The practice has a name: wash trading. One diner, rejoining the line all day, making the restaurant look busy from the street.

It gets worse wherever volume itself is rewarded. Some venues pay token rewards per dollar traded, which quietly inverts the metric: the reward funds the laps that farm the reward. Traders run circles through the venue not because they want the service but because the venue pays them per lap. Ask the spine question: who is paying? The venue itself, to make its own line look long. The same mechanic drives a lot of behavior in points and airdrop farming, where users manufacture activity because activity is what gets rewarded.

The defense is simple and almost nobody applies it: read volume next to fees actually paid, never alone. Real traders pay fees as a cost of getting something they want. Wash volume tries to minimize fees, because fees are the cost of the lap. A venue reporting enormous volume and negligible fee revenue is telling you, in public, that its line rejoins itself.

What is FDV, and how is it different from market cap?

The fourth row needs two bars, because one number is hiding behind the other.

Market cap is price times circulating supply: the tokens actually out in the world today. FDV, fully diluted valuation, is price times everything that will ever exist, including every token still locked for the team, investors, and future rewards. When only a sliver of supply circulates, the two bars tell wildly different stories about the same token at the same price.

Here is the arithmetic in the shape you will actually meet it, with invented numbers. A pitch says: price 2 dollars, 50 million tokens circulating, one billion total supply, "only a 100 million market cap, cheap!" The seller led with the smallest number on the board. Do the multiplication the pitch skipped: 2 dollars times one billion tokens is a 2 billion dollar FDV, twenty times the number said out loud. Buying at this price means implicitly accepting that valuation.

And the remaining 950 million tokens are not hypothetical. They arrive on a published calendar, called unlocks, mostly landing on insiders and early investors who were paid in tokens at a fraction of today's price. Which sets up the analyst's question, the one the pitch is designed to keep you from asking: who buys the other 950 million as it unlocks, and at what price? A low-float token with a giant FDV is a small boat with a scheduled tsunami of supply behind it. Buyers of today's float are not early. Very often, they are the exit the unlocking sellers are waiting for.

Could the unlocked tokens politely never sell? They do not have to sell. But treating scheduled supply as if it will stay offshore is hope, not analysis. Price the tsunami; be pleasantly surprised if it misses. Supply schedules, vesting, and float are the heart of tokenomics, and no metric on the dashboard matters more for a token you are thinking of holding.

What are the three questions to ask any DeFi metric?

Compress everything above and the method fits on an index card. For any number anyone shows you:

1. What does it actually measure? TVL measures parked capital, not users. Volume measures transactions, not demand. APY measures a rate under assumptions someone chose. Market cap measures the float, not the token.
2. Who benefits from it being big? Every metric has a constituency. Protocols want big TVL, venues want big volume, sellers want small-sounding market caps and tall-sounding APYs. The number you are shown was chosen by someone.
3. What does it cost to fake? This is the ranking question. Wash volume is nearly free. Rented TVL costs emissions. But fees actually paid are expensive to fake, because faking them means genuinely paying them. Every dollar of fee revenue is a user who paid for a service: the receipt drawer, not the line outside.

That is why analysts keep landing on the same habit: whatever the pitch says, open the protocol's page on DefiLlama, look at TVL next to incentives, yield next to what pays it, volume next to fees, and market cap next to FDV and the unlock schedule. No single row decides anything. The rows contradict or confirm each other, and the contradiction is the finding.

So are all these numbers useless theater?

After four rows and their lies, torching the whole dashboard feels tempting. That reaction overshoots, and it is worth being precise about where the lens stops.

The numbers are instruments, not verdicts. A pilot does not throw out the altimeter because it cannot fly the plane; no single gauge was ever supposed to. Rising TVL without incentives, real fee revenue, a float schedule with no tsunami: read together, that is a story which is expensive to fake. Each gauge lies alone and speaks in a chorus.

Nor is the fix "trust the aggregator." DefiLlama aggregates honestly, but it reports what the chain shows, and the chain records rented TVL and washed volume as faithfully as the real kind. A clean window is still a window onto a street where the line can be staged.

The theater begins in exactly one place: when a single number is quoted alone, by someone selling. "A billion locked" is not analysis. It is the first word of a sentence the seller hopes you never finish.

If you want this reflex drilled rather than described, this article is adapted from a checkpoint of Your First 90 Days in DeFi, our free interactive course, where you decode a mock dashboard row by row. The first three checkpoints need no account.

Related questions

What is a good TVL for a DeFi protocol? There is no threshold that makes a protocol good. What matters is composition and trend: how much of the TVL is unincentivized, how long it has stayed, and whether it grew from deposits or just from rising token prices. Small and unpaid beats large and rented.

Is a high APY in DeFi a red flag? Not automatically, but it is a question demanding an answer. High yields fed by real fees or funding payments exist. High yields fed by a protocol printing its own token are a countdown. The number itself tells you nothing; what pays it tells you everything.

Why is FDV higher than market cap? Because most tokens launch with only a fraction of their supply circulating. Market cap prices the tokens that exist in the market today; FDV prices the total that will ever exist. The gap between them is future supply, scheduled to unlock onto the market over time.

Can TVL be faked? It can be manufactured. A protocol can pay emissions to rent deposits, and looping through composable protocols lets the same dollar appear in several TVLs at once. The chain records the deposits faithfully; the staging happens in why the money came.

What is the hardest DeFi metric to fake? Fees actually paid by users. Faking fee revenue means genuinely paying the fees, so the fake costs what it claims. That is why analysts read every other number next to fees: volume without fees is laps, and TVL without fees is a parking lot.

Where to go next

Four numbers decoded, three questions to ask, one row of receipts. You can now stand outside any protocol and read its line like a native: who is standing in it, who paid them to stand there, whether the same people rejoin it hourly.

Two natural next steps. The companion piece, where DeFi yield actually comes from, gives you the five sources behind every APY on the board. Then how to analyze a DeFi protocol in 20 minutes turns both skills into a repeatable checklist. Or learn the whole map in order with Your First 90 Days in DeFi, short interactive checkpoints taught with a security auditor's eye, starting below.