Fees, revenue, and what actually reaches the protocol

Fees versus revenue

Total fees are what users pay to trade: taker fees, borrow costs, and whatever else the venue charges. That number describes gross economic activity.

Revenue is the portion the protocol retains once it has paid out the participants who make the venue function — maker rebates, liquidity provider shares, referral payouts.

Both are legitimate metrics, but they answer different questions. Fees measure how much business is being done; revenue measures how much of it the protocol keeps. Quoting the first while implying the second is one of the more common ways DeFi metrics mislead.

Where the money goes on a perpetual DEX

The split varies by design and is worth checking before comparing venues.

Maker rebates

Many order book venues pay makers to quote. That payment comes out of taker fees and never reaches the protocol.

Liquidity providers

Pool-based venues route much of the fee income to the pool, since the pool is carrying trader profit and loss.

Insurance fund

A portion is usually retained against liquidation shortfalls. It is protocol-held but committed rather than distributable.

Treasury or token holders

What remains. On some venues this is most of the fee income, on others close to none.

Using these figures in valuation

Comparing a token’s fully diluted valuation to annualised revenue produces a multiple that can be compared across protocols — the DeFi equivalent of a price-to-sales ratio.

Two cautions apply. Annualising a short window assumes conditions persist, and crypto activity is cyclical enough that a figure annualised from a busy month can be several times a realistic run rate. And a multiple built on fees rather than revenue will look far cheaper than one built on revenue, so comparisons are only meaningful when both sides use the same basis.

Why revenue is the more honest metric

Fees can be inflated by incentivised trading, since rewarded volume pays fees on the way through. Revenue is harder to manufacture: paying users to generate fees that are then paid back out leaves nothing behind.

When a token incentive programme ends, fees usually fall sharply and revenue falls with them. A protocol whose revenue survives that transition has demand rather than a subsidy, and that distinction is visible in the data long before it is visible in the narrative.