The numbers look like institutional conviction. Hyperliquid's team sold $165 million of HYPE tokens since the December 2024 unlock. The protocol's "aid fund" bought back $364 million. Ratio: 2.2 to 1. Net: $199 million of fresh demand that didn't exist before the unlock.
Then you read the fine print. The team disposed of 87.8% of its unlocked allocation. The fund purchased at an average price of $37.10. The team sold at a blended average of $38.10. Current market value: approximately $54.80. The fund is sitting on a paper gain north of 47%.
Read those numbers in sequence. Then ask the question nobody on crypto Twitter wants to ask: who funded the bid?
Because in a decade of reading on-chain flows — from auditing ICO TokenSale contracts in Paris during the 2017 boom to watching liquidity evaporate at precise block heights during the Terra collapse — I've learned one rule that has never failed me. When the seller and the buyer share a balance sheet, the trade stops being a trade. It becomes a transfer. And transfers don't create value. They relocate it, and pocket the difference in perception.
This is not a buyback story. It's a balance sheet story. Let me walk you through the mechanics.
The Context: A Fair Launch Meets an Unfair Schedule
For those who need the setup: HYPE is the native asset of Hyperliquid, a high-performance Layer-1 blockchain built around a central limit order book DEX. Total supply sits at approximately 1 billion tokens. The unusual detail is the allocation. Team share: 0.493%. That's roughly 4.93 million tokens. Most Layer-1s hand insiders 15% to 20% of the supply. Hyperliquid gave its own people less than half a percent. The internet crowned it the fairest launch in crypto.
That fairness narrative carried the token through its early months. But narratives don't feed unlock schedules.
Since December 2024, the vesting contract has been releasing approximately 540,000 HYPE per month to current and former team members. Over eight months, that totals the full 4.93 million unlocked tranche. The data, compiled by MLM monitoring, covers this entire post-unlock window through late July 2025. It reveals a two-sided flow: relentless team distribution against an organized repurchase effort from an entity labeled the "aid fund" — a treasury vehicle nominally responsible for ecosystem support that has, in practice, become the token's largest market participant.
Here's the ledger. On the sell side: 1.19 million tokens sold on public markets at an average price of $27.30, generating $32.5 million. Another 3.14 million tokens sold through OTC desks at $42.00, generating $132 million. Combined: 4.33 million tokens for $165 million, at a blended average of $38.10.
On the buy side: the aid fund accumulated 9.8 million HYPE at an average entry of $37.10, deploying $364 million. Monthly pace: approximately 1.23 million tokens, or $46 million.
On paper, this is a textbook support operation. In practice, it's more complicated. And the complications are where the edge hides.
The Core: Reading the Order Flow
Let's start with the OTC premium. The team sold public market tokens at $27.30 and OTC tokens at $42.00. That is a 54% spread between two tranches of the identical asset, sold within months of each other. In an efficient market, that gap should not exist. When I ran delta-neutral basis strategies after the 2024 Bitcoin ETF approvals, the spread between spot and futures rarely exceeded a few dozen basis points before arbitrageurs closed it. A 54% premium is not an inefficiency. It's a different asset class wearing the same ticker.
What did OTC buyers receive for that premium? Size without slippage. Discretion. No order book footprint. And in crypto, the absence of a footprint is a valuable asset. The unanswered question is whether those OTC counterparties were long-term institutions or intermediaries who now hold 3.14 million tokens searching for a final home. I watched this exact pattern during DeFi Summer in 2020, when I actively managed yield positions across Compound and Uniswap. OTC desks were the first signal of a whale preparing to exit. Tokens left through private channels, sat in cold wallets for 60 to 90 days, and then appeared on exchange books as "organic selling pressure." The mechanics are identical today. OTC is stealth distribution for institutional-sized exits.
Second, velocity. The fund commits roughly $46 million per month. The team sells $20.6 million per month. The coverage ratio is 2.28 to 1. Overlay those lines and the picture sharpens: the fund isn't merely absorbing team supply. It's buying an additional 1.3 times the team's monthly sell volume from the open market.
That is not investment. That is price maintenance. I have built these models. Every buyback program follows one of two mandates: seek return, or seek stability. The fund's behavior is the latter, without exception. And stability operations have a critical flaw. They are only sustainable while capital lasts.
Third, the runway. The fund has deployed $364 million. At the observed monthly velocity of $46 million, that represents roughly 7.9 months of support. Eight months have already elapsed since the unlock began. The arithmetic is blunt: the buyback operation is either in its terminal phase, or the fund controls additional capital that the public cannot see. Either way, you are trading against a counterparty with an undisclosed budget. In options terminology, that is undefined risk on the other side of your position. You don't know where their stop-loss lives, or whether they have one.
The averages also conceal a chronology. The $27.30 public sales happened earlier, when the market was still digesting supply and price discovery was fragile. The $42.00 OTC sales came later, after the fund had established its bid and the market had stabilized. That sequencing suggests the fund's presence created the very conditions that allowed the team to command a higher OTC price. The "premium" buyers paid was, in part, the fund's doing. That's not a market finding its level. That's a market being engineered toward one.
Fourth, the price structure. Current market value sits near $54.80. The team's blended sell price was $38.10. The fund's average buyback price was $37.10. The market is currently paying 44% more than the weighted average of every insider transaction in the past eight months.
Let that sink in. The people who understand the project best sold at $38. The market is buying at $55. Information asymmetry is not a theory in this market. It's a price chart.
Fifth, the composition of the sell. Eighty-seven point eight percent of the unlocked team allocation is already distributed. There are two conventional readings. The bearish reading: insiders are signaling, with nine figures of real money, that $38 is a fair price and they want out. The bullish reading: the majority of the unlock overhang is spent, and forward supply pressure is materially lower than the market fears.
There is a third reading that nobody quotes. The team sold 87.8% within the first eight months of unlock. That is not gradual diversification. That is an exit in force. When I audited ICO contracts during the 2017 boom, I learned to distinguish between founders who respected their lockups and founders who designed gates that opened like doors. The HYPE unlock opened on schedule. The team walked through it. And they were nearly all the way out before the market finished debating the signal.
Sixth, the fund's forward behavior. The fund entered at $37.10. The market is at $54.80. That is a 47.7% unrealized gain on a 9.8 million token position. At a certain point, that gain becomes a liquidation event. If the fund's mandate is supportive, selling is forbidden. If the fund's mandate is financial, selling is inevitable. When trust rests on the name "aid fund," but behavior mirrors a market operations desk, I trust behavior. The fund has been buying for eight months. Its balance sheet is finite. The most important chart in this entire story is not the HYPE price. It's the fund's wallet balance. And nobody has it on their screen.
The Contrarian Angle: The Backstop Is the Next Whale
The conventional read of this data is bullish. Fund buys 2.2 times what the team sells. Net pressure: positive. Retail sees a backstop. Smart money sees a backstop with a deadline.
The contrarian read is darker. The support buyer becomes the next seller.
Consider the worst plausible scenario. The aid fund holds 9.8 million tokens with a 47.7% cushion. The team's sell pressure is mostly exhausted at 87.8% distributed. The marginal buyer of the past eight months disappears. And in its place, the market confronts a new whale with a vested interest in managing its own exit before the next unlock narrative arrives. The market loses its largest buyer and gains a new seller in the same transaction. That is not a support operation. That is a relay race with a single runner.
The second layer is the self-dealing question. The fund's capital source is unconfirmed. If the fund was seeded from treasury holdings or a project-controlled wallet, then the $364 million buyback is economically identical to the team selling tokens to itself. The reported "net $199 million buy pressure" was never external demand. It was internal asset relocation, engineered to keep the price elevated while insiders reduced exposure. This is the oldest trick in the market-maker playbook, dressed in blockchain transparency.
Arbitrage doesn't create value. It exposes mispricing. The mispricing here is the public's willingness to treat a buyback as an external vote of confidence when it may be an internal transfer between two pockets wearing different jackets. The market is paying $54.80 for a token whose most visible buyer is spending money that may have originated from the seller on the other side of the table.
There is also a scale problem worth naming. The total supply is one billion tokens. The team sold 0.433% of that supply. The fund repurchased 0.98%. These figures are rounding errors in a market cap measured in tens of billions. The dollar amounts make headlines, but the actual supply impact is marginal. What matters is signal. Insider distribution at 87.8% of the unlocked tranche is the clearest conviction indicator available in this market. No buyback, however well funded, can rewrite that message. The bull market is masking it right now. Bull markets always mask technical flaws. That's what they're for. But the flows are already telling you what the price will eventually confirm.
The Takeaway: Track the Address, Not the Tweet
The bull case is simple. The fund keeps buying, burns the tokens, and the supply picture tightens. The bear case is simpler. The buyback ends, the fund's 9.8 million tokens become a fresh overhang, and the quarterly unlock cycle keeps feeding the market. The difference between those outcomes is visible on-chain before it is visible in price. You just have to know where to look.
Three signals. First, the aid fund's balance. Declining balance means the support runway is closing. Flat balance means the buyback pace will slow. Inflow to the fund means asking a harder question: where is the capital coming from, and what is the community being asked to repay?
Second, the OTC receiving wallets. The 3.14 million tokens that changed hands at $42 went somewhere. If they begin migrating to exchange addresses within 60 days, the stealth distribution has begun.
Third, the unlock schedule. The team has sold 87.8% of the current tranche. The next quarterly unlock is already scheduled. If it is larger than the current run rate, the season is not over. It is halftime.
Price levels matter. Current: $54.80. If price breaks $44, the OTC buyers are underwater and their exit accelerates. If price breaks $37, the fund's entire average position is under water. That is the floor. If the fund's address starts moving tokens to exchanges, the floor disappears entirely. That is your exit signal. Not before. Not after.
Risk isn't a number on a screen. It's a counterparty waiting for you to blink. Someone is holding 9.8 million tokens at a 47.7% profit. At some point, they won't be holding anymore. Options don't reward conviction. They reward precision. And precision means refusing to buy a headline. It means reading the balance sheet instead. Terra's code was poetry; Luna's exit was prose. The blockchains are different. The pattern isn't.
The gap between belief and reality is where the trade lives. This time, both sides of the trade are visible on-chain. That is rare. Use it.