Bitmine Immersion Technologies, Inc. BMNR

Key data

The business

BitMine started as a bitcoin operation, running immersion-cooled miners and hosting rigs for others. In mid-2025 it became something entirely different: an Ethereum treasury, chaired by Tom Lee of Fundstrat, whose one job is to accumulate ether and grow the amount of it standing behind each share. It raised billions in fresh equity, bought ether with it, and now holds a stack that dwarfs everything else on its books. The reported revenue, $46.5M in the May quarter at an 83% gross margin, is mostly the staking yield the company earns for helping run the Ethereum network; against a $14B value it is a rounding error. Nobody owns this for the revenue. The only edge a vehicle like this has is cost of capital: when the stock trades above the value of its coins, it can sell shares dear and buy ether accretively, and being the biggest makes that cheaper. Below that line the edge disappears, because anyone can buy ether directly. The company publishes its holdings in a weekly 8-K, which is the number that actually matters.

The numbers

Revenue went vertical as the stack and its staking grew, but the earnings line is noise. The multi-billion GAAP losses below are non-cash marks, not money leaving the building: operating cash burn in FY2025 was just $4.1M.

The pattern to read is the revenue ramp against the collapsing EPS: as they issued shares by the hundreds of millions to buy ether, the per-share losses swung wildly on marks while staking income climbed. Net income for the May quarter is not broken out in this pack; the operating loss was $11.9M, back near cash reality after two quarters of billion-dollar accounting swings.

FY2025's $349M "profit" and FY2026's ≈$9B "loss" are the same thing, mark-to-market on the crypto and related instruments, pointing in opposite directions. Ignore both. The whole compounding story lives in one place: ether-per-share. When the shares trade above the coins' value, every share sold to buy ether lifts ether-per-share; when they trade below it, as now, issuing shares quietly destroys it. At last filing book value was about $12.0B; the stack and share count have both grown since. What this memo believes that the tape does not: at roughly one times assets the vehicle has lost the premium that was its only real advantage, and the single print that settles it is the weekly holdings update read against the share count.

Management

Chairman Tom Lee is the face and the reason the raises got done. The insider ledger is otherwise empty of signal: two tiny purchases by Lori Love totaling about $15,400 in the last year, no sales, plan status not disclosed. The record that counts is capital allocation, and it is stark: the diluted share count roughly doubled, from about 326M last November to about 569M now, all of it to buy ether. Above NAV that was the smartest thing they could do. At or below NAV it takes ether away from existing holders. The one genuinely reassuring fact is the balance sheet: debt is effectively zero, debt-to-equity near 0.0001, so there is no lender who can force a sale at the bottom.

How it fails or surprises you

Ether falls hard. This is the dominant exposure and no filing resolves it. A 50% drop in ether roughly halves the NAV and the stock. The protection is not a hedge, it is the absence of debt: a drawdown here is one you sit through, not a wipeout, because no margin call exists.

The discount flips back to a premium (right tail). BitMine trades near or below the value of its coins today. If sentiment returns the way it did for the bitcoin treasuries, the stock re-rates above NAV and gains twice, from ether rising and from the multiple expanding, and accretive issuance restarts. The tell would be the weekly mNAV crossing back above one alongside rising ether-per-share.

They keep issuing below NAV. The fact the numbers explain least is the doubling of shares against a stagnant stock. If management treats capital raising as a reflex rather than a NAV-dependent choice, ether-per-share bleeds even as ether rises. The print that proves the read wrong: more shares and less ether-per-share on the next 10-Q.

Closing thoughts

The share count and ether-per-share on each filing settle whether management is destroying value, but the stock's level otherwise tracks Ether's price times whatever premium or discount the market grants. The distribution is therefore ether's distribution, wide and roughly two-sided, with a management question overlaid. The fatter near-term tail is a deep ether drawdown while the discount lingers, but the zero-debt balance sheet turns that into a survivable slump rather than a permanent loss, which is the whole reason to prefer this structure to a levered one. Against it sits a real right tail: ether rallies while the discount closes, and you are paid on both.

The bet is still that Ethereum rises and that this is a clean, unlevered way to own a large stack of it. What breaks it is not ether but management: a market price stuck below the value of the coins while the company keeps printing shares to buy more, quietly shrinking the ether behind each one you hold. Watch two numbers together, the share count and ether-per-share on the next filing. If shares rise and ether-per-share falls, the discount is earned and this is not an opportunity, it is a leak.