Metric catalog

The US-dollar value of all assets deposited in a pool, protocol or chain at a point in time.

Unit

US dollars

Data source

DeFiLlama /pools, schema-validated per pool before display.

Interpretation

An order of magnitude: how much capital is currently committed to this mechanism.

Limitations

TVL contains two moving parts — quantity and price. A 40 % rise can be pure price effect with not one additional dollar deposited. Double counting across stacked protocols is possible, and deposited capital is not tradable depth.

Precision

TVL is a stock measure at a point in time — not a flow, not an earnings figure, and not a measure of economic utility, solvency, tradable depth or security posture. Nothing follows from a TVL reading about whether the deposits could be withdrawn under stress. For lending markets, DeFiLlama reports either the total supply or the supply minus borrows, depending on the protocol (compared against /lendBorrow on 27 September 2026); across protocols the TVL there is therefore not one consistent quantity, and never the withdrawable liquidity.

Risk

Read as a quality signal, TVL misleads: capital follows incentives and leaves just as fast.

Institutional relevance

Sets the ceiling on what position size is conceivable at all without substantial price impact.

See this metric on real data →Glossary entry →EVD-2026-0008 →

The return projected to one year including compounding, derived from the currently observed rate.

Unit

percent per year

Data source

DeFiLlama /pools; depending on the pool as a total APY and split into base and reward components.

Interpretation

A snapshot, not a commitment: the rate is projected forward as if it stayed constant for a year.

Limitations

The underlying rate is variable and moves with utilization and the emission schedule. The quoted APY accounts for neither transaction costs nor impermanent loss nor taxes.

Precision

Five quantities have to be held apart. (1) Instantaneous rate: the rate in force at this moment. (2) Annualized rate (APR): the per-period rate r scaled linearly to a year, r × n, without compounding. (3) APY: the same rate with compounding, (1 + APR/n)^n − 1. (4) Realized return: what a given position actually produced, in retrospect. (5) Projected return: an assumption about the future. What is displayed is (3) alone, as DeFiLlama reports it — this platform does not compute it from a rate itself. The compounding formula presumes an r that stays constant across the year — a premise a variable market does not deliver. An APY on this platform is therefore neither (4) nor (5).

Risk

An unusually high APY signals that a question is open — which one, the number does not say.

Institutional relevance

Without the base/reward split, the APY is not usable for revenue planning.

See this metric on real data →Glossary entry →EVD-2026-0008 →

The part of the yield that comes from use of the market itself — trading fees or borrowing interest.

Unit

percent per year

Data source

DeFiLlama /pools, field apyBase.

Interpretation

The part that would remain if every token emission ended tomorrow.

Limitations

The base APY is variable too: it depends on trading volume or borrowing demand and falls with them.

Precision

Base and reward APY are separately reported fields (apyBase, apyReward), not quantities formed by subtraction; the source defines the total APY as their sum. Only one of the three fields has to be present: where the total is missing, the flag-point calculation derives the reward share from apyBase and apyReward rather than reading the total as zero; no total APY is displayed then, only “–”. Where apyBase is missing instead, the base share is unknown rather than zero — no zero is substituted, and the data flag “incomplete data” adds a point.

Risk

A base APY near zero alongside a high total APY means practically all of the return comes from emissions.

Institutional relevance

The repeatable component — the only part a revenue forecast can rest on.

See this metric on real data →Glossary entry →EVD-2026-0008 →

The part of the yield paid out in additionally distributed tokens.

Unit

percent per year

Data source

DeFiLlama /pools, field apyReward.

Interpretation

An incentive payment by the protocol, valued at the distributed token's current price.

Limitations

The valuation assumes the reward token is sellable at the price used. Where the reward token's liquidity is thin, that is exactly the open question.

Risk

Emissions end by schedule or by governance decision; the component can fall to zero with nothing else in the market changing.

Institutional relevance

Yield in a token the organization may not hold is operationally a forced-sale requirement.

See this metric on real data →Glossary entry →EVD-2026-0008 →

The change in a pool's APY against its level 30 days earlier, stated in percentage points (pp) rather than as a percent change: from 4 % to 6 % is +2 pp.

Unit

percentage points (pp)

Data source

DeFiLlama /pools, field apyPct30D, taken as reported. Where the value is missing, “–” is shown. The report assembles a market's path from this field together with the 1-day and 7-day fields.

Interpretation

Shows the direction and size of the latest move: whether today's APY is above or below its level a month ago.

Limitations

Two points in time are compared; the figure does not show what happened in between. An unusual value at either end determines it entirely.

Precision

This platform reads the field as a difference in percentage points, as the report and the glossary state it; this reading has not been independently checked against the source's own description.

Risk

A sharp fall can point to expiring incentives, a sharp rise to temporary scarcity or new rewards — the figure does not say which.

Institutional relevance

Shows whether today's snapshot is typical of the past month before it is projected forward.

See this metric on real data →Glossary entry →

The average of the daily APY values DeFiLlama reported over the last 30 days.

Unit

percent per year

Data source

DeFiLlama /pools, field apyMean30d, taken as reported; this platform does not compute the average itself.

Interpretation

Puts the current APY in context: if it sits well above the average, the snapshot is unusual for the past month.

Limitations

An average smooths — spikes and drops disappear into it. It describes the past and says nothing about future yield.

Precision

How the source treats days without a value and pools younger than 30 days has not been checked here; such an average may rest on fewer than 30 observations.

Risk

An average carried by a few very high days makes a pool look steadier than it was.

Institutional relevance

A calmer reference than a single day's value when yields are compared across markets.

See this metric on real data →Glossary entry →

The share of capital supplied to a lending market that is actually borrowed.

Unit

percent

Data source

Available for lending markets in the Explorer, and cross-checked against Aave and Morpho where a match exists.

Interpretation

Drives the rate: high utilization means high demand and, as a rule, higher rates on both sides.

Limitations

Very high utilization also means little free capital is available for withdrawals.

Precision

The metric is defined for lending markets and has no counterpart in an AMM pool, where “supplied” is not a borrowable quantity. The denominator is capital supplied, that is free balance plus amount borrowed; how a protocol assigns reserves, the reserve factor and non-withdrawable balances differs, so two protocols in the same state can report different utilization figures.

Risk

At near-full utilization a withdrawal can be effectively blocked until loans are repaid or liquidated.

Institutional relevance

The metric where liquidity risk in a lending market becomes visible first.

See this metric on real data →EVD-2026-0008 →

A lending market's supply minus the borrowed part — the most a withdrawal could find at this moment, arithmetically.

Unit

US dollars

Data source

Computed by this platform from the totalSupplyUsd and totalBorrowUsd fields of DeFiLlama /lendBorrow — only for markets for which that endpoint reports both values.

Interpretation

How much capital is free for withdrawals. The higher the utilization, the smaller the amount.

Limitations

A snapshot: every new loan and every repayment changes it. Protocol-specific caps and reserves are not taken into account.

Precision

The floor at zero hides a contradiction in the data — reported borrowing above supply — rather than presenting it as a liquidity shortfall these data cannot evidence.

Risk

Where it is small, a withdrawal may have to wait until loans are repaid or liquidated.

Institutional relevance

The ceiling on what position could be withdrawn again at short notice.

See this metric on real data →Glossary entry →

The rate, projected to one year, that borrowers in a lending market pay on borrowed funds.

Unit

percent per year

Data source

DeFiLlama /lendBorrow, field apyBaseBorrow. Reward discounts on the borrowing side (apyRewardBorrow) are not included; for markets that cannot be borrowed from, the value stays empty.

Interpretation

The cost on the demand side. Next to the supply APY it shows the spread between the borrowing and the deposit rate.

Limitations

Variable and driven by utilization. Without the reward discounts it is not the net rate a borrower actually bears.

Risk

At high utilization it can rise abruptly; a position built on borrowed capital then costs more than planned.

Institutional relevance

The basis of every calculation that involves borrowed capital, such as carry or leveraged positions.

See this metric on real data →Glossary entry →

The highest share of collateral value that may be borrowed against.

Unit

percent

Data source

Market parameters of the respective lending market, shown in the Explorer.

Interpretation

The highest permitted loan-to-value when taking out or increasing a loan — a borrowing limit, not a realization limit.

Limitations

The parameter is a governance decision and can be changed, including for existing positions. It says nothing about the ratio at which a position becomes liquidatable; this platform carries only the max LTV field, no liquidation threshold value.

Precision

Max LTV and the liquidation threshold are two separate market parameters and must not be equated. Max LTV caps how much may be drawn at all; the liquidation threshold — as a rule the higher of the two — sets the ratio at which a position becomes realizable. The distance to forced realization follows from the liquidation threshold and is measured by the health factor, not by max LTV. Alongside them sit further parameters in their own right: liquidation bonus, liquidation penalty, close factor, and borrow and supply caps.

Risk

A high permitted LTV moves risk into the collateral's price volatility.

Institutional relevance

Functionally equivalent to a haircut — only changeable unilaterally by governance rather than negotiated.

See this metric on real data →Glossary entry →EVD-2026-0008 →

The loan-to-collateral ratio at or beyond which a position may be closed by force by a third party.

Unit

percent

Data source

A market parameter of the respective lending market, to be evidenced from its documentation or contract state — not part of this platform's yields endpoint.

Interpretation

The point at which the market's mechanics turn against the position: below it the position is unremarkable, above it a third party may realize it.

Limitations

The threshold is a governance parameter per collateral asset and market, not a market price — it can be changed, and the same collateral carries different thresholds in two markets. This platform does not report the value.

Precision

Not to be confused with max LTV: max LTV caps borrowing, the liquidation threshold governs realization. Whether realization actually triggers depends in addition on the oracle price rather than the observed market price, and how much of a position may be closed in one step is set by the close factor.

Risk

The narrower the gap between max LTV and the liquidation threshold, the less price movement separates a permitted position from forced realization.

Institutional relevance

The threshold against which an internal limit or margin process would have to be defined — not the borrowing cap.

The ratio of weighted collateral value to outstanding loan; below 1 the position is liquidatable.

Unit

a ratio, dimensionless (threshold at 1)

Data source

Per position from the lending market's contract state; this platform looks at markets rather than individual positions and therefore does not carry the value.

Interpretation

The actual measure of distance to forced realization — the quantity mistakenly attributed to max LTV.

Limitations

A health factor is a snapshot of one position. It moves with the collateral price, with accrued interest and with any change to the liquidation threshold.

Precision

The price entering the formula is the market's oracle price, not an exchange's market price; the two can diverge. The weighting entering it is the liquidation threshold, not max LTV — substituting max LTV understates the figure and makes the position look closer to realization than it is.

Risk

A reading just above 1 is not a buffer: a small move in the collateral price can push it under.

Institutional relevance

The quantity position monitoring attaches to — and the basis of any margin rule.

A dispersion measure of a pool's observed APY values over recent history.

Unit

The same unit as the underlying APY series — percentage points where the APY is measured in percent. A standard deviation is dimensionless only where the series itself is.

Data source

DeFiLlama /pools, field sigma.

Interpretation

How reliable the current APY is as an expectation: low sigma means a stable rate, high sigma means today's value says little about tomorrow.

Limitations

A dispersion measure describes the past. A pool with a short history necessarily has a sigma that says little. On top of that, the estimator's parameters are not known here (see Precision) — without them the figure is a comparison between pools, not an absolute dispersion value.

Precision

A standard deviation is fully specified only once six things are stated: the observation window (lookback), the sampling frequency, the number of observations, whether the estimator is the population or the sample form (denominator n or n−1), the handling of missing values, and the handling of outliers. The DeFiLlama endpoint serves the sigma field as a finished number and documents none of those six; this platform does not compute the value itself and therefore cannot supply them either. That is an open question, not a parameter quietly filled in — which is also why the Explorer renders the value as a bare number carrying no unit symbol.

Risk

Low volatility of a rate says nothing about the failure risk of the underlying mechanism.

Institutional relevance

Relevant to the question of what range a revenue forecast should carry.

See this metric on real data →Glossary entry →EVD-2026-0008 →

DeFiLlama's flag for whether a pool carries impermanent-loss risk by its classification — yes, no or not stated.

Unit

flag (yes / no / not stated)

Data source

DeFiLlama /pools, field ilRisk. Where it is missing, this platform shows “–” rather than “No”; the data-flag points count the gap like a yes (factor 3).

Interpretation

Separates pools whose value depends only on the price of the deposit from those whose value also depends on the price ratio of two assets.

Limitations

A yes or no, not a size: the flag does not say how large a possible loss would be.

Precision

Not a measure of a loss that has occurred. Impermanent loss is realized only on withdrawal at a shifted price ratio, and fee income can offset it — the flag accounts for neither.

Risk

A “no” on a pool with several non-stable assets contradicts the pool's mechanics and should be checked against them.

Institutional relevance

Decides whether a position can be valued through one asset's price alone or through the ratio of several.

See this metric on real data →Glossary entry →

Whether a position is exposed to the price risk of a single asset or of several at once.

Unit

flag (single / multi / not stated)

Data source

DeFiLlama /pools, field exposure, taken as reported. Where it is missing, the data-flag points count the pool like one with several assets (factor 3).

Interpretation

With several assets, the risk that their ratio shifts comes on top of each asset's own price risk.

Limitations

Says how many assets are involved, not which: two stablecoins are just as “multi” as two highly volatile assets.

Risk

A “single” does not rule out that the one asset is itself a bundle, such as a liquid-staking or LP token.

Institutional relevance

Determines which prices a position would have to be hedged against or monitored for.

See this metric on real data →Glossary entry →

DeFiLlama's flag for pools whose reported APY lies so far outside the usual range that a data error or a short-lived incentive program is more likely than a lasting yield.

Unit

flag (yes / no)

Data source

DeFiLlama /pools, field outlier, taken as reported. This platform does not re-check the procedure by which the source sets it. A flagged pool receives a data flag (+2) and is not part of the report's curated universe.

Interpretation

A cue not to read the reported APY as a yield until it is clear where it comes from.

Limitations

A flag without a measure: it does not say how far the APY deviates, or from what. A pool without the flag is not plausible for that reason.

Risk

A single outlier can dominate averages, totals and rankings if it is not filtered out.

Institutional relevance

A filter to apply before any comparison across yields.

See this metric on real data →Glossary entry →

The value of assets traded through a market over a period.

Unit

US dollars per period

Data source

DeFiLlama /pools, fields volumeUsd1d and volumeUsd7d — only for pools for which DeFiLlama reports a volume, mainly DEX pools. For every other market, to be sourced from a DEX data source or directly on-chain.

Interpretation

Actual use, as opposed to merely deposited capital.

Limitations

Volume is manipulable: wash trading and an incentive program's own churn produce real on-chain volume with no economic substance.

Risk

Volume without fee income suggests the measured trading earns the protocol nothing.

Institutional relevance

The reference for how many days of trading a planned position corresponds to.

See this metric on real data →Glossary entry →

The total amount users of a protocol pay over a period.

Unit

US dollars per period

Data source

Not contained in this platform's pool data; to be evidenced from protocol documentation or a fee data source.

Interpretation

What the market is actually willing to pay for the service.

Limitations

Fees are not revenue: most of them regularly flow to liquidity providers, not to the protocol.

Risk

Treating fees as revenue systematically overstates a protocol's economic viability.

Institutional relevance

The starting figure for any valuation not based on token prices alone.

The share of fees retained by the protocol or its treasury.

Unit

US dollars per period

Data source

Protocol documentation and on-chain treasury addresses; not part of this platform.

Interpretation

What remains to the protocol itself — the figure from which it funds development and security.

Limitations

Revenue at the protocol is not return to the token holder; whether any mechanism connects the two is a separate question (dimension 5).

Risk

A protocol whose incentive spending persistently exceeds revenue is financing itself through dilution.

Institutional relevance

The only basis for an earnings-based valuation.

The number of distinct addresses that interacted with a protocol over a period.

Unit

count

Data source

On-chain indexing; not part of this platform.

Interpretation

A proxy for usage — explicitly not for user count.

Limitations

Addresses are not people. One person can hold hundreds of addresses, and an airdrop incentive produces exactly that.

Risk

Read as a user count, the metric substantially overstates adoption.

Institutional relevance

Meaningful only in relation to volume or TVL, never on its own.

Circulating token supply multiplied by the current price.

Unit

US dollars

Data source

A price and supply data source; not part of this platform.

Interpretation

A theoretical aggregate market valuation: circulating supply valued at the current reference price.

Limitations

The price applies to the last unit traded, not to all of them. The metric implies the whole supply is liquidatable at that price.

Precision

The metric is expressly not the amount that could be traded or liquidated today: a sell order that is small relative to market depth can already move the price substantially, and the price applies only to the last unit traded. Two further inputs are the data provider's conventions rather than measurements — what counts as “circulating” (the treatment of locked, treasury-held and burned tokens) and which reference price is used. Two providers can therefore report different market caps for the same token on the same day.

Risk

Where liquidity is thin, market cap can be a multiple of what is actually realizable.

Institutional relevance

To be used only together with trading depth and FDV.

Maximum token supply multiplied by the current price.

Unit

US dollars

Data source

The protocol's token contract and emission schedule.

Interpretation

What today's valuation would imply if every token ever scheduled were already circulating.

Limitations

FDV ignores timing: an unlock six years out weighs differently from one six weeks out.

Precision

The formula presumes a maximum supply is defined at all. Under uncapped inflation, a dynamic supply, an emission schedule governance can change, or simply an unknown future supply, that quantity does not exist, and an FDV is then either undefined or quietly computed from today's total supply — which is a different metric. Before comparing two FDV figures, establish which supply each provider used.

Risk

A large gap between market cap and FDV shows how much dilution is still ahead.

Institutional relevance

Relevant to whether the intended holding period crosses an unlock date.

Share of circulating supply that is staked and therefore not freely tradable in the short term.

Unit

percent

Data source

The staking contracts of the respective network or protocol.

Interpretation

A measure of how much supply is taken off the market by lock-up.

Limitations

Liquid staking tokens make the lock-up tradable — the ratio then no longer measures withdrawn supply.

Risk

Unbonding periods bite exactly when many want to exit at once.

Institutional relevance

Determines how quickly a staked position can be unwound at all.

Forced realization of posted collateral when a position falls below its threshold.

Unit

US dollars per period

Data source

The lending market's on-chain liquidation events.

Interpretation

The mechanism that keeps a lending market solvent — ordinarily a sign that risk control works.

Limitations

The figure alone does not distinguish orderly unwinding from a cascade.

Precision

Three quantities are frequently conflated: the volume of collateral realized, the number of liquidation events and the number of positions affected. One position can be realized in several steps, because the close factor limits what share may be closed at once — event counts therefore routinely exceed position counts.

Risk

Liquidations are procyclical: selling pressure lowers the price, which triggers further liquidations.

Institutional relevance

The central transmission path from market risk into credit risk.

Loan value left without sufficient collateral backing after realization.

Unit

US dollars

Data source

Protocol reports and the lending market's on-chain state.

Interpretation

Evidence that liquidation did not work in time, or not at sufficient prices, when it mattered.

Limitations

Bad debt is often quantified only after the fact; a zero reading does not mean none arose.

Precision

The formula given is a teaching simplification, not a protocol's loss accounting. A defensible measurement takes in at least: interest accrued up to realization, actual realization proceeds, the liquidation bonus and the costs of realizing, the oracle price used for valuation, the valuation of collateral still held, reserve coverage and insurance funds, and the protocol's own loss allocation including socialized losses. Which of these items a protocol recognizes, and how, is protocol-specific and must be evidenced from its own accounting.

Risk

Residual losses are distributed, depending on the protocol, to depositors, reserves or token holders.

Institutional relevance

The first question a bank asks: who bears the residual loss?