DeFiLlama’s metric definitions separate three points in a protocol’s cash-flow chain. Protocol fees are all amounts users pay to use a protocol; protocol revenue is the subset retained after payments to liquidity providers and other supply-side participants; holder revenue is protocol revenue returned through distributions, staking rewards, buybacks or burns.
The distinction matters because a DeFi token captures value only when a defined, enforceable route connects protocol cash flows to tokenholders. High fees can show user payments without showing that the protocol retains them or that holders receive them.
Confusing these measures turns platform activity into an unsupported claim about token value. The question is therefore not simply how much users pay, but how much is retained and what ultimately reaches tokenholders.
Protocol fees, protocol revenue and holder revenue
These terms describe different points in one economic chain. Treating them as interchangeable can turn a measure of platform activity into an unsupported claim about token value.
Protocol fees are gross user charges. On a decentralized exchange, they can include trading fees paid as swaps are executed. On a lending market, the relevant charges may arise from borrowing or other protocol activity. The label says nothing, by itself, about who owns the proceeds once they are collected.
Protocol revenue is what remains with the protocol, its treasury, team or tokenholders after the parties supplying a service have been paid. That distinction matters especially where liquidity providers, validators or other supply-side participants have an economic entitlement to most of the fees. A protocol may facilitate substantial activity and report substantial fees while retaining little or none of them.
Holder revenue is the final and most demanding category. It exists when retained revenue is actually returned to tokenholders through an identified mechanism. Funds sitting in a treasury may support the ecosystem, but treasury accumulation is not automatically tokenholder value capture. The appropriate sequence is therefore: gross fees, retained revenue, then the holder-level claim.
Trace where each user fee goes
The practical task is to follow a unit of fee income from the user to its eventual recipient. Start with the activity that produces the charge, identify the recipient specified by the protocol’s rules, then determine whether any part is retained and whether the retained amount has a mandated use.
Users are the source of fees. Liquidity providers and other supply-side participants may be the first recipients because they provide the capital or service that makes the protocol usable. The protocol treasury may receive a separate share, but only if the code and configuration direct funds there. Tokenholders receive value only if another rule or decision carries that treasury or protocol share onward.
That produces a useful audit trail:
- What fee does the user pay?
- Which participants receive it first?
- What fraction, if any, is retained by the protocol?
- Who controls the retained funds?
- Is a portion automatically distributed, used for buybacks or burns, or paid to eligible stakers?
A headline number at the first step cannot answer the later questions. Nor does the word “revenue” settle the issue if the protocol’s accounting convention stops at treasury retention rather than holder distribution.
Fee switches and protocol retention
A fee switch illustrates why a protocol’s capacity to collect revenue differs from actually collecting it. The switch determines whether part of user fees goes to the protocol instead of being allocated entirely to liquidity providers.
Uniswap v2 provides a clear design example. Its whitepaper specifies a 0.05% protocol fee when the fee switch is activated, equal to one-sixth of the standard 0.30% trading fee. Without activation, the existence of a trading fee does not mean that the protocol receives that 0.05% share. The fee paid by traders and the retained protocol fee are separate quantities.
The Uniswap v2 whitepaper therefore shows two separate questions that should not be collapsed: whether the protocol can redirect a share of fees to itself, and whether any retained share creates a claim for the token. Turning on a fee switch may create protocol revenue; it does not, on its own, specify a payment or benefit for holders.
This distinction also prevents a common analytical shortcut. A proposed switch, a dormant switch or a configurable parameter is not the same thing as current revenue. The operative settings and their recipients matter more than the theoretical maximum fee the design could collect.
How holders can receive value
Different accrual mechanisms create different economic relationships between a protocol and its token. Direct distributions and staking rewards can return protocol revenue to tokenholders.
Buybacks use protocol revenue to purchase tokens in the market, converting revenue into token demand rather than making a cash distribution. Token burns can also return value to tokenholders: burning purchased tokens reduces the relevant token supply, while retaining them can produce a different economic effect.
Research on DeFi valuation finds that value capture requires a defined and enforceable path from protocol cash flows to tokenholders, including fee sharing, buybacks, burns or staking distributions. Governance rights alone do not establish a revenue claim, and revenue held in a treasury does not automatically constitute value captured by tokenholders.
Automatic, discretionary or proposed accrual
Automatic, discretionary and proposed accrual can look similar in a token narrative, but they do not give holders the same claim. Rules specified in advance make an allocation automatic. Governance, a foundation, a treasury manager or another decision-maker can make a discretionary allocation, while a proposal signals possible future policy rather than a current entitlement. The distinction affects the expected timing, amount and recipient.
In the Aave discussion of a formal surplus-allocation framework, governance materials describe buybacks as a route from protocol revenue to demand for AAVE. They also note that governance can direct treasury funds toward growth, safety or other initiatives.
That range of choices can be rational for a protocol, yet it leaves retained revenue outside holder revenue unless a defined and enforceable path exists—for example, buybacks, distributions, burns or staking rewards. The practical test is whether the governing rules say what the system must do, rather than what a governing body may choose to do later.
Recurring revenue and event-driven fees
Even a valid accrual mechanism does not make every dollar of fee income equally durable. Revenue quality depends on the underlying activity and on whether the source recurs through normal operations or appears mainly in exceptional market conditions.
Liquidation fees are a prominent example for lending protocols. They can rise sharply during market dislocations, when borrowers are more likely to be liquidated. Such income may be economically real, but a spike does not demonstrate that the protocol can sustain the same level through ordinary conditions.
Aave DAO funding materials explicitly caution against annualizing liquidation revenue as sustainable income because it can surge during dislocations. The same principle applies more broadly: separate recurring operating fees from episodic sources before using revenue to assess a token’s prospective accrual.
A simple comparison can help. Suppose one protocol retains fees from regular user activity month after month, while another reports a similar total after a brief stress event generates large liquidation fees. Their reported revenue may match for a period, but the composition is different. Any analysis of a holder distribution, buyback capacity or burn rate should account for that difference rather than extrapolating the aggregate number.
Why revenue multiples are not enough
Research examining protocol revenue multiples reports large dispersion and a weak connection between protocol revenue and token valuation. It therefore considers revenue alongside token supply, potential dilution, distribution policy, sustainability, and the legal or governance enforceability of the value-accrual mechanism.
Revenue can be a useful input, but a revenue multiple cannot substitute for claim analysis. Gross protocol fees are the least direct measure; retained revenue is closer; and revenue returned to tokenholders through a defined mechanism is closer still. Treasury-held revenue does not automatically benefit tokenholders.
A buyback or distribution can have a different per-token effect depending on the number of tokens entitled to it, including future tokens that may enter circulation. A defined and enforceable path from protocol cash flows to tokenholders is distinct from governance rights or a stated intention, and a treasury may fund operating needs before any holder-directed action.
Frequently Asked Questions
Are protocol fees the same as protocol revenue?
No. Fees are the total charges paid by users, while revenue is the portion retained by the protocol after payments to liquidity providers or other supply-side participants.
Does a protocol treasury make its token revenue-backed?
Not automatically. Treasury assets become a tokenholder value-capture mechanism only when rules or enforceable decisions direct value from those assets or protocol revenue to holders.
What does a DeFi fee switch do?
A switch can redirect part of fees that would otherwise go entirely to supply-side participants toward the protocol, but it does not by itself determine how the protocol’s share will be used.
Are buybacks the same as direct tokenholder distributions?
They are different mechanisms: a distribution transfers value directly to eligible holders, whereas a buyback uses funds to purchase tokens. The effect of a buyback depends partly on how those tokens are treated.
Why should liquidation revenue be treated cautiously?
Liquidations can generate unusually high fees during market stress. That income should be separated from recurring operating revenue rather than assumed to persist at the same rate.
Can a low protocol revenue multiple prove a token is undervalued?
No. The multiple does not establish whether revenue reaches holders, whether the revenue is durable, or how token supply and future dilution affect any per-token claim.
