The DeFi Protocol That Generated More Revenue Than Goldman Sachs in a Single Quarter

One of the most legitimately unexpected data points in cryptocurrency last year came from the Hyperliquid narrative that surfaced through 2025. Running on its own specially designed Layer 1 blockchain, the decentralized perpetual futures protocol made about $844 million in revenue over the course of the year, surpassing Ethereum’s $524 million in protocol-level revenue during the same time frame.

Even experienced DeFi analysts were taken aback by the figure. The majority of the decentralized financial economy is hosted on the Ethereum network. The comparison was not intended to be feasible. The discussion of where value truly accumulates in the on-chain economy was reframed by the possibility that a single specialized protocol may outearn the network that underpins the majority of cryptocurrency’s smart-contract activity.

It is worthwhile to carefully examine the mechanics underlying the comparison. In this context, Ethereum’s revenue refers to the fees that the protocol itself collects, including the gas consumed through EIP-1559 and the larger fee structure that reimburses validators and takes ETH out of circulation.

As transaction volume that had previously settled directly on Ethereum’s foundation layer was absorbed by Layer 2 solutions, the number has been decreasing. 97% of the trading fees earned on Hyperliquid’s native exchange—perpetual futures, spot trading, and HLP vault transactions—are directed toward the buyback of HYPE tokens via what the protocol refers to as its Assistance Fund. There is a structural difference between the two sources of income. Even so, the comparison highlights a particular area where real economic activity has been concentrated.

The operational story of Hyperliquid is revealed by the trade volume figures that underlie its revenue. Throughout 2025, the platform handled almost $2.95 trillion in total trading volume, with peak months surpassing $400 billion. For comparison, Robinhood’s total cryptocurrency trading volume in 2025 was over $380 billion; during peak months, Hyperliquid momentarily outpaced Robinhood’s monthly volumes. In Q3 2025, the protocol’s market share in decentralized perpetual futures peaked at almost 80%.

That percentage had settled at about 44% by March 2026, but it was still significantly higher than rivals like Aster. A record for any DEX, the total percentage of worldwide perpetual contract trading (across controlled and decentralized venues combined) increased to over 6%. The growth was caused by real share shifting from centralized exchanges to Hyperliquid’s on-chain order book mechanism rather than a rising general tide.

For token holders, the revenue is especially significant because of the buyback flywheel. The automated daily HYPE buybacks from the open market receive about 97% of protocol fees. Higher trading volume generates higher fees, which finance larger buybacks, which lower the circulating supply, supporting the token price and drawing in more traders and liquidity. This process forms a self-reinforcing loop. By April 2026,

Hyperliquid had used this method to buy back and sequester about 44 million HYPE coins on its own. Additionally, in December 2025, validators authorized a permanent burn of 37.5 million HYPE tokens, which at the time were valued at about $912 million. This removed over 13% of the completely diluted supply from circulation. In a manner that most other cryptocurrency initiatives haven’t quite succeeded, the structural architecture directly connects exchange revenue to token economics.

More important than most media recognizes is the architectural distinction that lies beneath all of this. There was no pre-existing general-purpose chain upon which Hyperliquid was constructed. With sub-second block latency, entirely on-chain order book matching, 200,000 orders per second throughput, and zero gas fees on transactions, the team created a unique Layer 1 specifically tailored for trading.

Instead of using synthetic bridging, third-party DeFi apps can directly access native exchange liquidity through the HyperEVM smart contract layer, which is protected by the same HyperBFT consensus. The licensed S&P 500 perpetual contract, which surpassed $100 million in volume on its debut day in March 2026, was the result of the extension into real-world asset perpetuals using the HIP-3 framework, which enables builders to construct permissionless markets supported by conventional asset oracles. Currently, more than 30% of the platform’s total trading volumes are made up of TradFi assets.

Given what Hyperliquid’s 2025 figures truly indicate, there’s a sense that the larger story about which cryptocurrency networks are valuable is actually changing. In its most basic form, Ethereum was meant to be DeFi’s value-capture mechanism. The data from 2025 points to a more complex situation. With $1.3 to $1.4 billion in revenue,

Solana lead all blockchains, primarily due to its spot DEX volume and memecoin trading. Despite being a Layer 1 with only one core application, Hyperliquid came in second with over $844 million. The network that is said to be the foundation of DeFi, Ethereum, ranked fourth with about $524 million. By Q4 2025, Ethereum’s portion of all app income across all blockchains had dropped from around 50% in early 2024 to roughly 25%. The change illustrates a particular aspect of how value accumulates when application-specific chains persistently optimize for certain, high-value tasks.

Whether Hyperliquid continues on its present course through 2026 will depend on a number of unanswered factors, such as the repurchase flywheel’s endurance, emerging perp DEXs’ competitive response, and regulatory developments around decentralized derivatives. This is not advice on investments. Anyone attempting to comprehend how on-chain finance is actually developing in 2026 should, at the very least, pay close attention to this pattern.