BitMine Holds 4.8% of All Ethereum. Its Stack Is Still $5 Billion Underwater

BitMine Holds 4.8% of All Ethereum. Its Stack Is Still $5 Billion Underwater

One company now holds 5,847,611 Ether — 4.8% of every Ethereum token in existence, worth roughly $14 billion — and last week it added 32,447 more, about $81 million, in the middle of ETH's strongest weekly rally in more than a year. The buyer is BitMine Immersion Technologies (NYSE: BMNR), a company that spent fourteen months turning itself from a bitcoin miner into the world's largest corporate Ethereum treasury. The scale is the story the headline sells. The economics behind it are the story an investor needs.

BitMine, chaired by Fundstrat's Thomas Lee, began an "Ethereum treasury strategy" on June 30, 2025 with a stated target to own 5% of all ETH — it calls the goal the "Alchemy of 5%" and says it is 97% of the way there. About 87% of its holdings are staked through its own validator network, MAVAN, which the company says produces roughly $330 million of annualized staking income at current yields near 2.7%. With $14.9 billion in crypto and cash, it runs the second-largest digital-asset treasury behind Michael Saylor's Strategy, and it has bought ETH every week since the strategy began.

The 4.8% figure deserves to be broken apart, and the first number to pull out is the cost basis. Analyst-tracked data put BitMine's average purchase price near $3,366 per ETH. The company accumulated aggressively in late 2025, when it was paying roughly $3,800–$3,900, then kept buying down through ETH's slide from a $4,764 high to a $1,507 low. ETH now trades around $2,440–$2,480 — about a quarter below BitMine's average cost — which means that even after the best week in 14 months, the position remains roughly $5 billion under water. The company itself frames the rally as historical precedent, noting similar weekly gains preceded subsequent runs of roughly +170% and +167%; that is a sample size of two, quoted by the seller of the thesis. The hard account is more sober: the mark-to-market loss on the treasury was reported near $6.6 billion in February and $8.9 billion in June, and this month's surge has trimmed it to roughly $5.4 billion. Improvement, not profit.

The second number to pull out is what the money used to keep buying costs. Unlike many crypto-treasury companies, BitMine carries essentially no bank debt — about $30 million — so it is not one margin call away from being a forced seller. Instead it funds by issuing fresh equity and a 9.50% Series A perpetual preferred stock that began trading in June (an upsized offering that raised about $280 million). A perpetual preferred is a bond-like claim that stands ahead of common stock and pays a 9.5% cash dividend with no maturity date. Set the two rates side by side: the ETH itself yields about 2.7% staked, while the marginal capital used to buy more of it costs 9.5%. That roughly seven-point gap is a standing toll the asset cannot pay from its own income — it has to be covered by rising ETH prices, or it erodes common equity.

The equity-funded half carries a subtler cost: dilution. Every new share issued near a market low to buy more ETH leaves existing holders with a smaller slice of a position that may itself be underwater. The arithmetic shows up in both the stock and the accounts. BMNR has lost roughly 89% from its all-time high and 51% in the first half of 2026 even as its treasury nearly doubled in size. In its most recent reported quarter (ended May 31) the company posted $46.5 million of revenue, mostly staking income, and recorded an $83.6 million net loss attributable to common shareholders. The asset grows; the per-share claim does not necessarily follow.

Even the people running the machine conceded the leverage cuts both ways this summer. After raising its buyback authorization to $4 billion in April, the company spent about $86 million in a single July week buying its own stock — and slowed weekly ETH purchases to roughly $14 million in that same stretch. By early August it had repurchased about 16 million shares. Read that as a boardroom signal, not decoration: with the stock near $20 and ETH near $1,900, management preferred its own shares to more Ethereum. Only after ETH jumped 30% this month did BitMine swing back with its biggest weekly ETH purchase since early July. The "relentless weekly buyer" is actually a price-sensitive marginal buyer — which cuts against the popular notion that its accumulation acts as a permanent floor under the market.

The market has already priced most of this, which is the last thing worth examining. Near $24 a share, the company's market capitalization is roughly $14 billion — about equal to the value of the ETH alone. There is no discount embedded for the preferred stock sitting ahead of the common, none for the operating losses, and none for the risk of a retest. You are paying full price for the underlying asset while accepting the leverage's demonstrated downside, an explicit 9.5% carry cost on new money, and ongoing dilution. That is a package a retail investor can assemble more cheaply by owning ETH directly or through a regulated product, keeping the staking yield and dropping the single-entity risk.

Two clarifications keep the picture honest. One: 4.8% of supply is not "control" of Ethereum — not a governance majority and nowhere near a consensus threshold. What it is, is unprecedented concentration of supply and staking in a single public company, and that concentration is a market-structure fact in its own right: the largest standing demand engine for ETH is a company that must keep issuing securities to keep buying. Two: BitMine is roughly 188,000 ETH — at recent weekly pace, weeks — from reaching 5%. The milestone is close enough that the interesting question is what the company does the week after it arrives.

For a holder of BMNR, the whole thesis is the leverage itself: if ETH keeps climbing from roughly $2,400 toward the $3,300-plus average cost, billions of paper losses convert to gains quickly, and a stock down 89% can re-rate violently in the process. That is a legitimate bet, but it is a leveraged bet with a known toll, not ownership of a moat. The forward observables that would test it are concrete: whether ETH holds above roughly $2,400, whether the equity issuance that funds the purchases outruns the buyback program, and whether the near-$330 million of staking revenue covers the 9.5% preferred dividend and operating costs. The 4.8% headline is true. What it is worth depends entirely on which side of $3,366 the next stretch of Ethereum — and BitMine's share count — lands.