Fee switches, buybacks and burns are choices, not entitlements

A fee switch is a parameter that redirects part of a protocol's fee income toward token holders or a treasury, usually by burning units bought with that income or by distributing it. Every such mechanism exists because a governance process decided it should, and it persists only while that decision stands. Nothing about holding the token obliges anyone to keep paying.

Where the money is before anyone routes it

Fees paid by users are split before any of them reach a token holder. On a lending or exchange protocol, the largest share typically goes to whoever supplied the capital or liquidity that made the service possible; on a chain, to validators or miners who produced blocks. That share is supply-side revenue, and it is a cost of providing the service rather than a residual.

What remains after the supply side is paid is what a protocol can direct. The take rate is the proportion of total fees the protocol retains, and it is the parameter that governance actually controls. Holder revenue is the narrower figure that reaches token holders through a burn, a distribution, or an equivalent mechanism. These three quantities are frequently reported under the single word revenue, and the distinction changes the number by an order of magnitude in some cases.

Why it resembles an equity multiple and is not one

Placing market capitalization over annualized fees produces a figure that reads like a price-to-earnings ratio, and market cap to fees is a legitimate comparison as long as its differences from that ratio are held in view.

Fees are not earnings. They are gross payments by users, measured before the supply-side split and before any cost of running the protocol, contributors, audits or grants. Nothing is accrued, matched to a period or audited. A quarter of unusually heavy activity is recorded as it happens with no smoothing.

There is no residual claim. A shareholder owns a claim on what remains after creditors, and directors owe duties in exercising it. A token holder generally owns a governance right and nothing more. If holder revenue is zero because no fee switch is on, the fee figure in the denominator accrues to a treasury or to the supply side, and the holder's connection to it is a hope that a future vote redirects it.

The denominator can be paid in the token being valued. Where fees are collected in a protocol's own units, or where reported revenue includes incentives funded by issuance, the ratio is partly self-referential. Holder revenue yield and revenue yield invert the ratio into a percentage, which invites comparison with an interest rate; the same caveats apply, and the percentage form makes them easier to forget.

Buyback and burn against direct distribution

Two mechanisms move retained income to holders. A buyback spends income acquiring units, which are then destroyed by burn; every remaining holder's proportional share rises without anyone receiving anything. A distribution pays income to holders who meet a condition, often staking or locking, and the holder receives an asset.

The differences are practical. A buyback benefits passive holders automatically and requires no claim, no eligibility rule and no address list; a distribution requires an act by the recipient and can exclude those who do not perform it. A buyback's effect depends on the market it buys into, and buying with income the protocol itself generated is trading against the same market it serves. Tax and regulatory treatment of the two differs across jurisdictions and is not a matter this site can generalize about.

Implementation details are worth reading rather than assuming. A buyback executed by a contract that spreads purchases over time behaves differently from one executed at a moment by a discretionary operator, and whether the purchases happen on an order book or against an automated market maker determines who is on the other side. Some programs are authorized as a maximum rather than a commitment, so the announced size is a ceiling that need not be reached. And where a protocol simultaneously pays incentives funded by newly issued units, the buyback and the issuance partly cancel; reporting one without the other describes a gross flow as though it were a net one.

The legal shadow over both is worth stating without overstating. A mechanism that channels profits from an enterprise's efforts to passive holders is one of the factors courts consider under the Howey test in the United States, and comparable questions arise under other frameworks including MiCA, which entered application in the European Union in 2024. This is a live area, positions differ by jurisdiction and by facts, and the observable consequence is simply that some protocols with the technical ability to switch on fee routing have chosen not to. That is context for reading a switched-off fee switch, and it is not legal advice.

Reading a fee switch as a governance object

The verifiable questions are procedural. Whether the switch exists in deployed code or only in a proposal. Who can enable it, and whether that is a token vote, a multisignature wallet, or a foundation. What quorum a change requires, and whether concentrated voting power makes the outcome a formality. Whether a timelock separates a vote from its execution. Whether a previous switch has been changed, and how quickly.

A mechanism running for years with wide participation and a slow amendment path is a different object from one a small group can alter next week, even where the current parameters are identical. Recorded governance history is available in a way that intentions are not.

The fees pages separate total fees, supply-side revenue, protocol revenue and holder revenue for each asset that publishes enough to support the split, and valuation shows the ratios built on each line. The last lesson in this track covers the treasuries that hold retained income.