The gap between what users pay and what token holders receive sits at the center of Bitwise's latest CIO memo on crypto's growing revenue base. For investors, the useful question is no longer just whether people pay to use these products. It is what happens to the money once they do, and whether any of that activity actually reaches the token in a durable way.

Bitwise Says Crypto's Revenue Era Has Arrived

Bitwise's analysis focuses on an industry where large applications increasingly have measurable paying demand rather than relying entirely on token incentives and speculative activity.

The hard numbers come from Bitwise's Q2 research. The firm found that the 10 largest crypto applications generated a combined $5.9 billion in user fees over the 12 months through June 30, 2026. PancakeSwap, Hyperliquid, and Aave were the largest, each approaching $1 billion.

That activity held up through a difficult market. Bitwise estimates that crypto asset prices fell 36% during the first half of 2026, while major trading and lending applications continued producing substantial fee volume. That makes the fee chart useful not just as a revenue snapshot, but as a sign that some crypto products are being used in ways that persist even when prices are under pressure.

There is an important detail beneath the chart. Bitwise defines the $5.9 billion as total fees paid by users. It is not $5.9 billion of profit, nor $5.9 billion available to token holders.

Where the Money Goes After a User Pays

A trading fee can be split among liquidity providers, market makers, referral partners, or outside developers. A lending protocol may direct part of its economics into a DAO treasury. Another project may use revenue to buy its own token.

Gross fees indicate whether users are willing to pay for a product. Retained revenue reveals more about a protocol's economics. Buybacks, burns, distributions, or other mechanisms determine whether any of that activity creates an economic link to the token.

Those layers can look radically different even among the applications sitting near the top of Bitwise's chart.

Hyperliquid and Aave: Contrasting Revenue Models

Hyperliquid Turns Part of Its Fees Into HYPE Demand

Hyperliquid had passed $1.2 billion in cumulative trading fees by July, according to Grayscale data sourced from Allium. The number represents gross platform activity, not money flowing entirely to HYPE holders. Rebates, liquidity mechanisms, outside market deployers, and other parts of the system take their share.

What makes Hyperliquid unusual is what happens to part of the remainder. Fees allocated to the protocol's Assistance Fund are automatically converted into HYPE, and those acquired tokens are burned. Trading activity therefore creates recurring purchases of the asset while simultaneously reducing supply.

HYPE holders do not receive those fees as a dividend and do not gain an equity claim on Hyperliquid. The economic connection comes through market purchases and supply reduction.

Aave's Revenue Path Runs Through Governance

Aave generates its economics through borrowing and lending rather than perpetual-futures trading, and its relationship with AAVE is less mechanical.

The difference was visible in March, when the "Aave Will Win" proposal passed its Temp Check with 52.6% support. The framework sought to route revenue from Aave-branded products into the DAO treasury, including economics generated by Aave.com swaps, its mobile app, Aave Card, and Aave Pro. The proposal still had further governance stages ahead before becoming binding.

With Hyperliquid, part of the fee engine automatically reaches HYPE purchases. With Aave, decisions about where product economics belong and how they should ultimately benefit the ecosystem pass through governance first.

Stablecoins: A Completely Different Revenue Model

Some of crypto's largest revenue businesses barely resemble either model.

Tether and Circle issue dollar-backed tokens and invest the reserves supporting them in assets that can earn interest. Their economics can therefore expand even without charging users large trading fees.

Stablecoin issuers generated roughly $5 billion from their Ethereum deployments in 2025, with reserve income providing the core engine behind the sector's profitability. The same analysis put Tether and Circle's combined wider revenue at nearly $8 billion for the year.

The source of that money changes the risks around it. A derivatives exchange needs trading volume. Lending revenue responds to borrowing and leverage. Stablecoin reserve income depends heavily on circulating supply and interest rates.

A quiet market can hurt an exchange while leaving stablecoin balances intact. Falling rates can cut into reserve income even if stablecoin adoption continues to grow.

Revenue Quality: Durable or Fragile?

Fee totals say little about how expensive the activity was to attract. A protocol can pay large token incentives, rebates, or liquidity subsidies while reporting impressive gross fees. Another may generate the same amount with much less spending required to keep users active.

Cyclicality matters as well. Hyperliquid benefits from active markets and volatility. A prolonged drop in derivatives volume would slow the same fee engine that supports HYPE purchases. Lending applications face their own cycle as leverage expands and contracts.

The persistence of activity becomes more informative than one annualized number. Investors can examine whether fees survive weak markets, whether users return without heavy subsidies, and how much of the money remains after incentives and other payouts.

This is a tougher standard than total value locked. A large TVL figure shows that capital has entered a protocol. It does not show whether anyone will pay to use it.

High Revenue Does Not Determine Token Value

Revenue makes some crypto assets easier to analyze, but the comparison with public companies breaks down quickly.

Buying shares normally gives an investor legal ownership rights in a business. Holding a crypto token usually provides no automatic claim on profits, treasury assets, or residual value. Any economic benefit must come from the particular design of the protocol—perhaps a buyback, burn, staking mechanism, or governance-approved distribution.

The valuation question then becomes much narrower: how much economic value does the protocol capture, how durable is it, how much reaches the token, and how much growth is already reflected in the token's market value?

A protocol generating hundreds of millions in fees can still have an expensive token if investors have already priced in years of expansion. A smaller application can offer stronger economics relative to its valuation even with a lower absolute revenue figure.

Traditional multiples can help frame that discussion, but gross protocol fees cannot simply be dropped into a stock-style price-to-sales calculation and treated as equivalent corporate sales.

Crypto Is Getting a Harder Set of Fundamentals

For years, crypto projects could point to token prices, TVL, or user counts without demonstrating that customers would repeatedly pay for the product.

Large fee-generating applications raise the bar. Paid usage can now be measured, compared across cycles, and tested against the incentives required to produce it.

The next filter is stricter. A protocol has to show not only that money enters the system, but that the economics are durable and that the asset investors buy has a credible reason to benefit from them.

Revenue makes crypto easier to measure. It also makes weak token economics harder to hide.

Disclaimer: This article is for informational and educational purposes only and does not constitute financial or investment advice. Protocol fees, revenue, and token-value-accrual mechanisms can change through governance, market conditions, or protocol upgrades, and past revenue does not guarantee future token performance.