随着财政部不断购买，炒作鲸鱼转向交易所

A roughly $16 million exchange deposit and a separate $5.78 million sale have been combined into headlines about a $22 million whale dump. The distinction matters because Hyperliquid Strategies has been accumulating HYPE too. Its SEC filing shows how a public treasury chooses when to buy and what its 33.2 million-token position really represents.

Summary Lookonchain tracked 177,518 HYPE worth $16.08 million deposited to OKX and Bybit on September 28.

A separate wallet it linked to Hypersphere Ventures sold 62,869 HYPE for about $5.78 million.

A wallet linked to Hyperliquid Strategies bought 494,200 HYPE worth $45.8 million over 16 hours on September 25.

The company’s September filing reported approximately 33.2 million HYPE held as of September 8.

About 25.1 million of those tokens were delegated to its linked validator, with 8.1 million to Anchorage.

One HYPE holder sent 177,518 tokens, worth approximately $16.08 million at the time, to OKX and Bybit on September 28. A different wallet linked by Lookonchain to Hypersphere Ventures sold 62,869 HYPE for about $5.78 million. The tracker’s original item identifies one exchange deposit and one sale. It does not establish that the full $21.86 million was sold.

The distinction is the story. An exchange deposit makes coins readily available to trade, and it can precede a sale. It can also reflect custody, market making, collateral or a later withdrawal. A realized sale has a different evidentiary status. Headlines treating the combined $22 million as executed selling have already promoted a possibility into a fact.

On the other side, a wallet that Lookonchain linked to Hyperliquid Strategies acquired 494,200 HYPE worth about $45.8 million over 16 hours on September 25. That purchase was three days before the September 28 whale activity. Comparing their dollar amounts as if the treasury took those exact coins off the seller on the same day would be false. Public transfers do not name the other side of every trade.

One deposit and one sale were reported

The 177,518-token transfer went from a wallet identified as 0xc745 to two centralized exchanges. The 62,869-token sale was attributed by the tracker to a different wallet it associated with Hypersphere Ventures. The latter transaction was estimated to realize roughly $2.13 million in profit. Wallet attribution is a researcher’s classification, not a statement signed by the institution, so the association should be kept attached to its source.

The $16.08 million deposit and $5.78 million sale sum to $21.86 million at the prices used in the report. That is the basis for the rounded $22 million figure. Only $5.78 million is described in the underlying tracker account as an executed sale. The other $16.08 million is a transfer to venues where trading could happen later. Even that valuation can change before the tokens are sold.

Suppose all 177,518 deposited HYPE eventually sold. The combined disposed amount would be 240,387 HYPE including the separate 62,869 sale, subject to transaction details and potential double counting. Suppose none of the deposit sold. The confirmed amount remains 62,869 HYPE under the tracker report. A responsible article should not choose either endpoint without following the exchange balances and execution evidence.

Transfers into an exchange omnibus address make the next step harder to trace. Once coins are credited internally, exchange trades occur on the venue’s private ledger. An outside analyst may observe deposits, subsequent withdrawals or order-book activity, but cannot usually match a precise deposited token with a buyer’s identity. The absence of a visible onchain sale after deposit does not prove the holder kept the position. Nor does a deposit itself prove disposal.

This is why the phrase head for the exits is too strong as a factual headline for the 177,518 HYPE movement. A holder moved inventory into tradable venues; another sold a smaller, documented position. The two events together show potential supply pressure, not a single $22 million completed dump.

The treasury’s purchase is real, but from a different window

The September 25 crypto.news report described 494,200 HYPE bought by a wallet linked to Hyperliquid Strategies for about $45.8 million. The tracker attributed the activity to wallet 0x6436. Buying across a 16-hour period is an observed action from a labeled address, not a corporate filing for that exact intraday amount.

At the stated prices, the treasury purchase was about 2.85 times the later $16.08 million exchange deposit and about 7.9 times the separate $5.78 million sale. Those ratios compare public dollar figures from different dates; they do not show one party absorbed the other’s inventory. The September 25 buy may already have been complete before September 28’s transfer. They belong in a chronology, not a netted same-day order book.

The company’s formal filings provide a broader check. In its September registration statement, Hyperliquid Strategies reported approximately 33.2 million HYPE held as of September 8. It said the balance included about 12.5 million tokens contributed at the transaction closing and tokens purchased with proceeds from a roughly $299.9 million PIPE and other capital raises. A 494,200-token reported acquisition later in September is about 1.49% of that September 8 inventory. That is meaningful incremental accumulation, but the dated 33.2 million figure cannot be treated as a current balance after it.

JUST IN: Hyperliquid open interest climbed to $14.3 billion

The print is the latest snapshot of positioning on the perps venue pic.twitter.com/fTGu9O89ic — crypto.news (@cryptodotnews) September 9, 2026

The company says it updates its HYPE balance weekly with a one-week delay. That disclosure lag matters when comparing a fast whale headline with treasury holdings. A balance published today might describe last week’s ownership. A look at an attributed wallet can fill part of the gap, but it may omit other wallets, internal transfers or custodial accounts. The filing and the wallet tracker are complementary evidence with different scopes.

The public company’s stock, PURR, provides exposure to HYPE through a corporate structure. Shareholders do not own a redeemable fraction of the onchain wallet. The issuer can buy more tokens, stake them, raise equity, repurchase its own shares and maintain cash reserves. The share count and obligations determine HYPE per share. A headline about a large treasury buy does not by itself prove accretion for an existing shareholder.

A filing spells out when the company intends to buy

Hyperliquid Strategies’ registration statement describes a Treasury Committee composed of its CEO, CFO and COO, subject to board oversight. It says the company seeks to raise equity primarily when its shares trade at a premium to a market net asset value measure. It then considers buying HYPE when the market price is below an internal assessment of long-term fundamental value. It may repurchase its own stock at a meaningful discount to net asset value.

Those are the company’s stated policies, not a guarantee that each purchase meets an external definition of cheap. The internal fundamental-value estimate is not an observable market price, and the filing does not disclose a public threshold at which the September 25 buying was approved. A reader can test the subsequent outcome through HYPE per diluted share, cash, issuance prices and actual purchases, but cannot reverse-engineer the committee’s complete judgment from an address label.

The same filing describes a committed equity facility of up to $2.5 billion with Chardan. Capacity is not cash already raised, and an equity facility can issue new shares. If future share sales finance HYPE purchases at attractive premiums, HYPE per share can rise; if shares are sold too cheaply or expenses absorb proceeds, it can fall. The company’s objective is to maximize long-term HYPE exposure per common share. The measurable unit is therefore per share, not the gross treasury balance.

You might also like: D3 offers priority access to new domains via Solana, Hyperliquid vaults

The capital framework separates a working capital reserve intended to cover at least 12 months of projected operating expenses, a cash reserve available for deployment and HYPE holdings. This explains why a company can announce a large cash raise without immediately buying an equivalent amount of tokens. Some cash is reserved for the business, and the committee can wait. It also means that its buying may continue through a market decline without any promise to support a particular HYPE price.

Another detail weakens the simple whale versus treasury picture. The company’s own filing says HYPE sales are conditioned on its assessment that the prevailing price exceeds its estimate of fundamental value. It is an accumulating treasury, but its policy does not forbid sales. A change in market price, cash requirements or its valuation assessment could put the company on the other side of the market. Present buying as an observed period, not a perpetual floor.

The 33.2 million tokens are largely staked

As of September 8, the company said approximately 25.1 million HYPE, or 75.7% of its staked holdings, were delegated to its linked HSIxUNIT validator. Another approximately 8.1 million, or 24.3%, were delegated to Anchorage, described as its only third-party validator at that time. The two balances sum to 33.2 million, matching the rounded treasury figure. The percentages refer to its staked position, not the entire circulating HYPE supply.

The staking detail changes the liquidity interpretation. HYPE moved to a validator is not available for immediate spot sale until the owner initiates unstaking and completes the applicable process. The filing describes an initial one-day lock after delegation and a seven-day unstaking queue following withdrawal from staking. That does not make the treasury immovable; it sets a timing constraint. The company can keep cash and use other financing while coins are delegated.

Staking rewards also enlarge holdings without a new market purchase. The filing cited an average net annualized reward rate of 2.18% at 440.4 million HYPE staked across the network as of September 8. Applying 2.18% to 33.2 million tokens gives about 723,760 HYPE annually as a rough run rate, before the company’s specific fees and changes in balance or network rate. It is not a forecast of actual annual rewards. Rewards accrue and compound under the protocol mechanics, while the market price can change the dollar result substantially.

If a future report shows HYPE holdings rising by 700,000 tokens, it would be wrong to call all of that new spot buying without separating staking rewards, transfers and acquisitions. Conversely, a flat balance could conceal purchases offset by sales or other uses. The split between purchased tokens, contributed inventory and earned rewards is essential to testing whether the treasury is absorbing outside supply.

The initial 12.5 million HYPE contribution is another reason to keep categories straight. Those coins became company property through its corporate transaction, not through an open-market buy on the date the company reported 33.2 million. Roughly 20.7 million tokens above that initial contribution existed in the September 8 balance, but even that difference cannot be labeled net spot purchases without accounting for staking rewards and other transactions since closing.

Buybacks respond to trading revenue, not whale headlines

Hyperliquid’s protocol routes most eligible trading fees through its Assistance Fund to purchase HYPE. Tokens acquired under the program are burned according to the protocol’s published structure. That creates an ongoing source of token demand tied to platform activity. It is separate from Hyperliquid Strategies, the public company, even though both can acquire the same asset.

The reported percentage of fees directed to the fund varies across descriptions because fee categories and program terms matter. One crypto.news unlock analysis used 99% of eligible trading fees; another account described 97% of protocol trading fees. The precise percentage should be quoted with its denominator and date. For the present comparison, what matters is the mechanism: trading generates fees, which fund market purchases, while actual buyback size changes with volume and pricing.

An example shows why a constant percentage is not a price guarantee. If eligible fees were $1 million in one day and 99% flowed into the fund, the buyback budget would be $990,000 before execution considerations. A $16.08 million exchange deposit would be more than 16 times that illustrative daily amount. If fees rose tenfold, the relative scale would change. Without matching a real daily fee tally and dated onchain fund transactions, one should not claim the mechanism absorbed September 28’s specific deposit.