Hunting Ethereum Bulls: "Whales" Suffer $70 Billion in Losses, Under Collective Scrutiny

Tom Lee and Jack Yi probably haven't been able to sleep well these past couple of days.

After all, if we were to pick the most dramatic protagonist for the crypto market in early 2026, it likely wouldn't be Bitcoin, nor some newly emerged narrative, but rather these two ETH "whales" who have been publicly put on the hot seat.

Onlookers never think a funeral is too big of a spectacle.

These past few days, global investors have been holding their breath, collectively watching how the largest and most transparent batch of openly declared long positions in Ethereum's history are struggling to survive amidst unrealized losses.

1. ETH "Whales" Are Already Facing Tens of Billions in Unrealized Losses

The stories are similar year after year, but the "whales" are different each time.

The term "whale," in the Web3 context, typically refers to institutions or individuals with substantial capital, enough to influence market trends.

But in recent years, the positive connotation of this word has been continuously diluted by reality, morphing into not just a heavyweight presence, but more like the most conspicuous and easily targeted spectacle during periods of severe market volatility.

Over the past few days, the two most discussed ETH "whales" in the market have undoubtedly been BitMine (BMNR.M), led by Tom Lee, and Trend Research, helmed by Jack Yi. Although both are Ethereum (ETH) bulls, they represent two entirely different paths: the former is the treasury company holding the most ETH, while the latter is an investment firm openly adding leverage on-chain and publicly going long on ETH.

First, look at BitMine.

As one of the most representative Ethereum reserve companies, BitMine once ambitiously proposed a long-term goal to acquire approximately 5% of Ethereum's total supply. As of the time of writing, the company has accumulated holdings of 4,285,125 ETH, with a market value nearing $10 billion.

According to ultra sound money statistics, the current total Ethereum supply is about 121.4 million ETH. This means BitMine has directly locked up roughly 3.52% of the circulating ETH supply. The pace of realizing this vision is undeniably aggressive.

It's important to note that it was only after completing a $250 million private placement in July 2025 that BitMine officially began its "Ethereum treasury" transformation. In other words, in less than half a year, BitMine completed the leap from a Bitcoin mining company to the world's largest ETH holder.

Source: ultra sound money

What draws even more attention is that even during last week's darkest hour when ETH broke below $3,000 and the market accelerated its collapse, BitMine chose to add to its position against the trend, buying an additional 41,787 ETH (approximately $108 million) at around $2,601, demonstrating steadfast conviction.

But the problem then arises—cost. Conviction has a price. BitMine's average ETH holding cost is approximately $3,837. This means that after ETH retreated to around $2,350, its paper losses have expanded to about $6.4 billion.

This extremely aggressive "coin-based" transformation has also triggered an exceptionally frenzied valuation game in the US stock secondary market.

Looking back to July 2025, when BitMine first began disclosing its Ethereum purchase strategy, its stock price (BMNR.M) was hovering around $4. Subsequently, the stock price completed a leap from the floor to the heavens within half a year, reaching a high of $161, becoming the most dazzling "Ethereum shadow stock" in global capital markets.

However, what rises with Ethereum, falls with Ethereum. As ETH prices experienced a deep correction, the premium bubble in BitMine's stock price rapidly burst, and it has now plummeted all the way to $22.8.

If BitMine represents the long-term spot path of trading time for space, then Jack Yi's Trend Research has chosen another, significantly riskier road.

Since November 2025, Trend Research has been openly and explicitly going long on ETH on-chain. Its core strategy is a typical "staking-borrowing-buying-restaking-borrowing" loop:

• Stake held ETH into the on-chain lending protocol Aave;
• Borrow the stablecoin USDT;
• Use the USDT to buy more ETH;
• Continuously cycle to amplify long exposure;

The actual logic of this operation is not complicated; it essentially uses existing ETH as collateral to borrow funds to buy more ETH, betting on leveraged gains when prices rise.

This is undoubtedly a highly potent strategy in a favorable market trend, but the risk stems precisely from this. Once the ETH price falls, the collateral value shrinks, and the lending protocol will require additional margin. Otherwise, it will trigger forced liquidation, selling ETH at market price to repay the debt.

Therefore, when ETH plummeted from around $3,000 to a low of approximately $2,150 within just 5 days, this mechanism was forced into a "stress state." The chain then presented a dramatic spectacle of "death by a thousand cuts":

To prevent the position from being forcibly liquidated, Trend Research continuously transferred ETH to exchanges, sold it for USDT, and then deposited the USDT back into Aave to repay loans, barely lowering the liquidation threshold and buying breathing room.

As of February 2nd, Trend Research has deposited a total of 73,588 ETH (worth approximately $169 million) in multiple batches to Binance for sale and loan repayment. The total loss on its ETH borrowing positions reached $613 million, including realized losses of $47.42 million and unrealized losses of $565 million. It currently still carries stablecoin leverage loans of about $897 million.

Especially during ETH's rapid descent through the $2,300-$2,150 range, the entire network was almost watching this "stop-loss survival" drama unfold in real-time—every ETH sold by Trend Research was both buying itself survival space and inadvertently handing new selling pressure to the market, further tightening the noose around its own neck.

In other words, Trend Research almost killed itself.

Source: Arkham

2. The "Ice and Fire" Contrast Between On-Chain and Off-Chain

Paradoxically, if we temporarily step away from the whales' tens of billions in unrealized losses and look at Ethereum from an on-chain structural perspective rather than price itself, we find a reality almost opposite to secondary market sentiment—ETH's on-chain activity is continuously heating up.

The Block's statistical data shows that approximately 36.6 million ETH are currently staked on the Ethereum Beacon Chain, exceeding 30% of the network's circulating supply and setting a new historical high.

It's important to note that the previous highest staking rate record was 29.54%, occurring in July 2025. This round marks the first time Ethereum's staking rate has substantially crossed the 30% threshold since entering the PoS era.

Source: The Block

From a financial supply-demand structure perspective, this change itself is highly significant.

A large amount of ETH being staked means it has voluntarily exited the free circulation market, undergoing a dramatic transformation from a "speculative currency" used for high-frequency trading and speculative games into a "yield-bearing bond" with productive properties. In other words, ETH is no longer just Gas, a medium of exchange, or a speculative tool. To a larger extent, it is beginning to play the role of a "means of production," participating in network operation through staking and continuously generating yield.

Of course, heavyweight players like BitMine, a treasury player, are integral to this—BitMine has already staked nearly 70% of its held ETH (approximately 2,897,459 ETH) and is continuously adding more.

Simultaneously, subtle changes have appeared in the validator queues. The staking exit queue is nearly empty, while the entry queue for staking continues to lengthen, with over 4.08 million ETH queued up waiting to "enter." In summary, the current state is: "Exiting is smooth, but entering requires a 7-day wait."

This queue size has already set a new high since the launch of Ethereum's PoS staking mechanism. From a time dimension, the steep rise of this curve precisely began in December 2025.

This also coincides with the beginning of Trend Research's openly aggressive long positioning on ETH.

Source: Ethereum Validator Queue

It's important to emphasize that, unlike trading behavior, staking is a low-liquidity, long-cycle allocation method that emphasizes stable returns. After all, once funds enter the staking queue, it means giving up the possibility of flexible portfolio adjustments and short-term speculation for a considerable period.

Therefore, as more and more ETH chooses to re-enter the staking system, it at least conveys a clear signal: at the current stage, an increasing number of participants are willing to bear the opportunity cost of long-term locking in exchange for deterministic on-chain yield.

Thus, a highly tense structural picture emerges. On one side, nearly one-third of ETH is continuously being "hoarded," with a steady stream of ETH waiting off-chain to be locked up. On the other side, secondary market liquidity is tightening, prices remain under pressure, and whales are forced to stop losses, with their positions frequently exposed.

This clear divergence between on-chain and off-chain constitutes the most vivid "ice and fire" landscape within the current Ethereum ecosystem.

3. The Openly Declared "Whale" Was Already on the Menu?

In traditional financial games, cards are often not openly transparent. Details like positions, cost bases, leverage ratios, etc., can be hidden within tools like derivatives and OTC protocols that benefit from information asymmetry.

But on-chain, every transfer, every collateralization, every liquidation line of a whale is exposed 24/7 to the view of the entire market. Once you choose to openly declare a long position, you easily fall into a physically draining war of "proving Murphy's Law."

So from a game theory perspective, although Tom Lee and Jack Yi are both bulls and have both declared their hands, they stand at opposite ends of the risk curve.

Despite facing $6.4 billion in unrealized losses, Tom Lee's BitMine has chosen the "low leverage, high staking, zero debt" spot path. As long as no structural risk is triggered, he can choose to wait it out within the time window, letting staking yield gradually hedge against volatility.

The facts are indeed so. Contrary to what many in the market imagine, BitMine's structure is not aggressive. As Tom Lee emphasized in a social media post on February 2nd: They have $586 million in cash reserves, and 67% of their ETH is staked, generating over $1 million in daily cash flow. For him, the decline is merely a shrinkage in paper numbers, not the onset of a survival crisis.

Jack Yi, however, has added leverage through Aave's looped lending, thus falling into the negative cycle of "price drop - approaching liquidation - transferring ETH to sell - adding margin - price drops again." It's like a "performance art" piece watched by the entire network.

Shorts don't necessarily need to liquidate you; they just need to depress the price → force position reduction → create passive selling pressure → trigger follow-on selling to complete a structural hunt.

This is why every loan repayment, every transfer by Trend Research is amplified and interpreted as a change in Jack Yi's confidence or whether he is about to capitulate. As of writing, Trend Research has stopped losses by selling 73,588 ETH (worth approximately $169 million), and the liquidation price for its borrowing positions has been lowered to below $1,800.

On the same day Tom Lee posted, Jack Yi also publicly reflected: As the person under the most pressure on the entire network right now, I must first admit that being bullish on ETH too early was a mistake... Currently, the profits from the previous round have been given back. While controlling risk, I continue to wait for the market to move upward.

Ultimately, going long via on-chain looped lending is equivalent to laying all your cards on the table for everyone to see. Whether or not there is an organized targeted attack, when you publicize your position, cost basis, leverage ratio, and liquidation line on-chain, you have already placed yourself on the hit list of all market resonance forces.

Of course, to some extent, this is also a form of path dependency. After all, in April 2025, Jack Yi publicly called for going long on ETH when it fell to $1,450, continuously adding to his position, and ultimately welcomed a rebound and profits, once becoming a "spiritual banner" for ETH bulls.

It's just that this time, the direction of the story remains unknown, and Tom Lee's chances of winning are clearly a bit higher.

In Conclusion

From Three Arrows Capital to FTX, to the now publicly watched BitMine, the script never changes: all collapses begin with excessive arrogance towards long-term certainty.

As Keynes's overused quote goes: "In the long run, we are all dead." Jack Yi's mistake is not in being long-term bullish on Ethereum, but in underestimating the cruelty of the market during short-term irrational phases. From the moment he chose to openly add leverage, he had already offered himself as a sacrifice to this transparent algorithmic world.

But looking at it from another angle, this might be the necessary "great cleansing" Ethereum must undergo. Every cycle needs such a whale fall process: whales being watched, leverage being squeezed out, path dependencies being shattered, and筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码筹码