forbes.com
$HYPE touched an all-time high above $62 on May 21, 2026, and the explanation on offer was the one markets always reach for. Institutional money had arrived. The first US spot Hyperliquid exchange-traded funds had begun trading days earlier, the token's fully diluted value briefly passed Solana's, and the financial press settled into a tidy story about Wall Street discovering a decentralized derivatives exchange.
That story is incomplete in a way that changes how the rally should be read. $HYPE is climbing in large part because Hyperliquid spends almost everything it earns buying $HYPE. The buyback is written into the protocol, it runs continuously, and it has very little to do with whether outside investors believe in the asset. The difference between treating the current price as a verdict and treating it as a mechanism comes down to understanding how that buyback works.
The Assistance Fund is the actual buyer
Hyperliquid runs a mechanism called the Assistance Fund. By DefiLlama's accounting, 99% of trading fees from the exchange's perpetual and spot markets flow into it, and the fund spends that money buying $HYPE on the open market. No board can vote to pause the program to conserve cash. The buying is the default behaviour of the protocol's revenue model, and it executes in every block and in every market condition.
The amounts are large enough to move the token on their own. Hyperliquid has produced more than $1.16 billion in cumulative revenue since launch, and effectively all of it has gone into acquiring its own token. In the third quarter of 2025 the protocol bought back $316.76 million of $HYPE across three months. Very few public companies return capital to shareholders at anything close to that intensity, and the ones that do make a deliberate decision each quarter to authorise it. Hyperliquid has removed the decision entirely. The buyback simply happens.
A second permanent bid sits underneath
The protocol is not the only programmed buyer of $HYPE. Hyperliquid Strategies, a company that listed on Nasdaq under the ticker PURR after a reverse merger with a former biotechnology firm, was built for one purpose, which is to accumulate and hold $HYPE. It owns roughly 20 million tokens. Its most recent quarter produced a $152.5 million net profit, and almost the entire figure was unrealised gains on the $HYPE already sitting on its balance sheet. A treasury company whose reported earnings rise and fall with the token it was created to hoard functions, in market terms, as a second standing bid, and that bid grows louder every time the price climbs.
A third stream runs through the stablecoin layer. When $USDC became Hyperliquid's aligned quote asset, the arrangement directed up to 90% of the reserve yield earned on $USDC held across the platform back to the protocol, again for buybacks and ecosystem incentives. Billions of dollars of $USDC sit in Hyperliquid accounts at any moment, so the interest earned on those balances alone generates another nine-figure annual flow pointed at the same token. Three separate pipes, then, all emptying into $HYPE.
The business underneath is genuinely strong
The buyback would recycle nothing if the underlying business were weak, so the fundamentals deserve their due. Hyperliquid has captured a dominant share of on-chain perpetual-futures trading, a category that has expanded quickly as traders look for venues outside the large centralised exchanges. Its cumulative perpetual volume runs into the trillions of dollars, and the fees feeding the Assistance Fund are real fees from real trading.
That last point separates Hyperliquid from a long list of earlier crypto projects, which manufactured the appearance of activity by paying users in their own inflating token. Hyperliquid earns close to a billion dollars a year from genuine customer activity, and a business that returns that money to token holders is doing something more honest than most of the industry has managed. The argument here is not that Hyperliquid is a fragile company. The argument is that the company and the token are different things, and the market is pricing the token as though they were identical.
What the ETFs actually contributed
Set the buyback machine beside the news hook the coverage led with. Bitwise launched the first US spot Hyperliquid ETF in May, and the early funds attracted tens of millions of dollars in their opening week. That is genuine institutional money, it is a credibility marker for a young asset, and it is worth having.
It is also modest in scale. A protocol buying back its own token at a rate of hundreds of millions of dollars a quarter operates an order of magnitude above an ETF inflow measured in tens of millions. The ETF launch became the headline because it fits a familiar template, the one where traditional finance validates a crypto asset. The Assistance Fund is the part actually setting the price, and it works the same way in weeks when no ETF story runs at all.
The two forces behave differently, and the difference is the point. ETF demand reflects choices by outside investors who can sell tomorrow. The buyback reflects an accounting consequence of people trading perpetual futures, and it would continue at full strength if every ETF holder lost interest overnight, as long as the trading volume held up.
What the buyback actually does for holders
The word buyback carries assumptions from equity markets that travel poorly to a token. When a public company buys back stock, it returns cash to shareholders by shrinking the share count, and that cash came from profit the company chose not to reinvest. A shareholder can sell into the buyback and walk away holding dollars.
The Assistance Fund hands nothing back. It converts protocol revenue into $HYPE that the protocol then holds, which removes tokens from circulation and supports the price. A $HYPE holder cannot redeem against the fund or claim a share of it. The value the fund creates is expressed through one channel only, the token's market price, and that market price is the very thing the buying is holding up. A $HYPE holder, in plain terms, owns an asset whose price is supported by a continuous bid, and the size of that bid is set quarter by quarter by trading volume.
The flywheel runs in both directions
That dependence is where the structure turns into a risk. A buyback funded by trading fees can never be larger than trading volume allows, and crypto trading volume is deeply cyclical. The protocol's own figures already show the effect. Quarterly buybacks fell from $316.76 million in the third quarter of 2025 to $255.05 million in the fourth, and then to $192.25 million in the first quarter of 2026. The standing bid under $HYPE shrank by roughly 40% across two quarters, during the same stretch in which the token went on to set record highs.
Price and engine moved in opposite directions, and that gap is the part the institutional-demand narrative leaves out. In a genuine crypto drawdown, perpetual-futures volume contracts hard, the buyback contracts with it, and the support fades at the exact moment $HYPE holders most want a buyer in the market. The mechanism amplifies gains on the way up and withdraws support on the way down. Only the first half of that cycle has been tested at scale.
The Solana comparison flatters the token
The claim that $HYPE passed Solana deserves the same scrutiny. The crossover happened on fully diluted value, the measure that prices in every token that will ever exist. On circulating market capitalisation, the measure of what is actually trading, $HYPE remains well below Solana, because a large share of $HYPE supply has not yet reached the market.
The unlock schedule matters for the buyback. As locked supply enters circulation, the Assistance Fund has to absorb steadily more potential selling simply to keep the price flat. The flywheel has to keep accelerating against a rising float, quarter after quarter. A slowdown in volume and an increase in circulating supply would arrive together, and they would compound.
How to price a token that buys itself
Hyperliquid remains one of the most profitable operations in crypto, with a revenue line most layer-one chains cannot approach, and routing that revenue into buybacks is a defensible way to reward holders. Arthur Hayes has put a $150 target on $HYPE by August, and given the mechanics, the target is internally consistent. Technical analysts watching momentum indicators have called the token overheated, which is also consistent with the mechanics.
The honest framing is that $HYPE's bull case and its bear case are the same sentence. The price is mechanically tied to Hyperliquid's trading volume, because the volume funds the buyback and the buyback funds the price. An investor buying $HYPE at a record high is taking a leveraged position on one variable, namely whether perpetual-futures volume on a single exchange keeps rising. That is a far narrower wager than a broad bet on decentralised finance, and narrower still than a bet on a general-purpose chain such as Solana. A chart at an all-time high looks like a market that has reached a conclusion. With this token, the market is mostly reading its own reflection.