BitMine (BMNR) 10-K FY2026: ETH Put Options Lost More Than Staking Earned
Listen to this report
The Hook
BitMine Immersion Technologies $BMNR spent fiscal 2026 turning what it calls the largest ether treasury into an income business, and the staking did its part: $121.4M of staking and validation revenue in the year to August 31, 2026, 97% of everything the company booked. Behind the $6.3B net loss that The Block led with on November 11 sits a second income strategy, the one the 10-K describes as "the sale of put options." That program lost $143.7M in the year, more than every dollar the staked ether earned. In the single quarter to May 31, it lost $92.1M against $46.5M of total revenue.
The filing's other plain sentence: "staking revenue was non-cash as it was received in digital assets." Of the $121.4M, $112.7M arrived as ether the company kept. The bills arrive in dollars: a 9.50% preferred dividend paid weekly, about $33.3M over the next 12 months, plus overhead, plus a buyback with $3.6B of authorization still open. All of it was paid for by selling stock, and on November 20 the stock closed, per stockanalysis.com, at about 0.73 times the ether behind each share.
Company Snapshot
BitMine, based in Norwalk, Connecticut and listed on the NYSE, was a small bitcoin miner until mid-2025, when it raised equity to buy ether and brought in Thomas Lee as chairman (Executive Chairman in this filing). This brief reads its 10-K for the fiscal year ended August 31, 2026, filed November 10, the first audited by KPMG; management again concluded that internal control over financial reporting was not effective. At year end it held 5,901,837 ETH with a fair value of $14,558.9M against a cost basis of $19,948.2M, about $3,380 per ether. It stakes most of it through MAVAN, the validator business built on its March 2026 acquisition of Pier Two, and reports one segment. The board has moved the fiscal year end to December 31, so the next annual document is a transition report for September to December 2026.
The Financial Story
The loss is the ether, and so is most of the offset. The net loss was $6,313.4M, of which $6,229.1M was the unrealized fall in digital assets: the filing's own figures put ether at about $4,406 a token at August 31, 2025 and $2,467 a year later, a 44% drop. Two non-operating lines cushioned it. The warrants sold with the September 2025 direct offering (exercise price $87.50, expiring March 2027) are a liability marked to the stock, so the stock's own fall produced a $257.9M gain. And tax: BitMine was charged $92.3M of deferred tax in fiscal 2025 on last year's unrealized gain, then booked a $92.3M benefit this year as the gain reversed. Cash payments for income taxes were zero in both years. The tax was charged and then un-charged; it was never paid.
The run-rate turned, quarter by quarter. The first half carried $298.4M of general and administrative expense, most of it in the quarter to November 30, 2025, which the 10-Q put down to "one time capital raising, advisory, legal, and other consulting fees." By the fourth quarter G&A was $41.2M and revenue was $64.9M. The table derives each period from the 10-K and the third-quarter 10-Q:
| FY2026 period | Revenue | Cost of sales | G&A | Before options | Loss on options | After options |
|---|---|---|---|---|---|---|
| Sep 2025 to Feb 2026 | $13.3M | $2.5M | $298.4M | -$287.5M | $41.2M | -$328.7M |
| Mar to May 2026 | $46.5M | $5.7M | $37.3M | $3.5M | $92.1M | -$88.6M |
| Jun to Aug 2026 | $64.9M | $10.1M | $41.2M | $13.6M | $10.4M | $3.2M |
| Full year | $124.8M | $18.3M | $376.9M | -$270.4M | $143.7M | -$414.1M |
The second-quarter 10-Q reported its own first-half figures, $65.3M of option losses and $298.6M of G&A; the third-quarter 10-Q's nine-month totals do not reconcile to them, so the first-half row is the one consistent with the later 10-Q and the 10-K's full year.
The fourth quarter is the first in which the business cleared its costs, before any change in the price of ether: $3.2M after the option losses. That is the part of the story the $6.3B net loss hides, and it is real.
The put program is the part the staking story hides. BitMine began writing ETH options in the quarter to February 28, 2026, "primarily through the sale of put options," as a way to earn more from its holdings. Premiums received over the year were $121.4M, almost exactly the staking revenue, but the company also paid $125.3M of premiums on options of its own, and by May 31 the 10-Q counted $79.3M of losses on options that were exercised, which for a put writer means buying ether above the market. The year closed with no open contracts, but the Business section still lists "ETH option premium income from option strategies on our ETH holdings from time to time" as a revenue source. In the March to May quarter, the options cost the company twice that quarter's entire revenue.
The income is in ether; the obligations are in dollars. Operating activities used $305.1M of cash in the year. The company sold almost none of its ether, $3.3M of digital assets in all, so everything was funded by equity: $13.2B of at-the-market stock sales, $365.2M from the September 2025 direct offering and $280.0M from the preferred. The preferred is expensive money. BitMine sold 3,500,000 shares at $80 against a $100 stated amount, so the 9.50% dividend is $9.50 a year on $80 of cash raised, an 11.9% cost, paid weekly, compounding toward 15% if unpaid, and redeemable at $100 on a change of control. The 10-K's own risk list names "liquidity constraints affecting dividend coverage." Year-end cash was $545.1M. The company's weekly 8-Ks then reported cash and marketable securities of $498M on November 1, $475M on November 8 and $448M on November 15, while it bought another 17,239 ETH over those two weeks.
The stock bought back at half the price it sold. The capital-return record, from Item 5 and the equity notes:
| FY2026 period | ATM shares sold | ATM gross proceeds | Avg sale price | Shares repurchased | Avg buyback price |
|---|---|---|---|---|---|
| Sep 2025 to May 2026 | 340,748,312 | $11,869.9M | $34.83 | 0 | n/a |
| Jun to Aug 2026 | 54,387,961 | $1,331.6M | $24.48 | 20,819,071 | $17.00 |
| Full year | 395,136,273 | $13,201.5M | $33.41 | 20,819,071 | $17.00 |
The repurchases ran 15,070,511 shares at $16.58 in July and 5,748,560 at $18.09 in August, $353.8M in all, leaving $3,646.2M of the $4.0B authorization. Shares outstanding went from 232,412,324 to 613,267,022 over the year, and the cover page counts 626,451,021 on November 4, 13.2 million more than at year end. Buying back below the value of the ether is accretive, but it needs cash, and the authorization left is about eight times the cash pile.
Valuation: what it's worth as a normal business
Where it trades. BitMine closed at $12.98 on November 20 (stockanalysis.com), $8.13B on the 626,451,021 cover-page shares. The November 16 8-K valued 6,077,745 ETH at $1,823, which is $11,080M, or $17.69 of ether per share: the stock is at about 0.73 times its ether. Adding the 8-K's other holdings (total $11.8B, including $448M of cash and securities and the Beast Industries and Eightco stakes) and deducting the $350M preferred and $33.7M of year-end liabilities gives net assets of about $11.4B, $18.22 a share, or 0.71 times. Strategy, the bitcoin treasury peer, showed about 1.09 times on its own dashboard on November 13. The two covering analysts average a $45.30 target (stockanalysis.com, November 21), and B. Riley cut its target to $27 from $34 around November 13 per the same site.
A DCF of the spread, plus the ether. A treasury is worth its net assets plus whatever its operations add or subtract. The operating spread here is staking income less the cost of running the company and the preferred dividend. Assumptions: the 8-K's $245M annualized staking at $1,823, scaled with the ether price; costs at the fourth-quarter run-rate of $205.5M a year (cost of sales plus G&A, stock compensation included), with the Ethereum Tower asset fee moving with the holdings; $33.3M of preferred dividends; the spread capitalized at the house 9% discount rate and 3% growth.
| Case | ETH price | Annual spread | Value of spread | Net assets | Implied value per share |
|---|---|---|---|---|---|
| Bear | $1,200 | -$68M | -$1.13B | $7.63B | $10.36 |
| Base | $1,823 | $6M | $0.10B | $11.41B | $18.38 |
| Bull | $2,466 | $83M | $1.38B | $15.32B | $26.66 |
At $1,823 the business roughly breaks even after the preferred, so the value is the ether. Re-rated to Strategy's 1.09 times, the base net assets would be worth about $19.86 a share; at par, $18.22. Under these assumptions today's $12.98 prices ether at roughly $1,400, or prices $1,823 ether with a permanent discount of about 29%. One reading of that gap is a charge for the things the table cannot see: an option program that can restart, fees that come off the top before common holders, and a funding model that only works above 1 times. Implied value under stated assumptions; not a price target and not investment advice.
Risks
Ether is nearly the whole balance sheet: each $100 move in the price is worth about $608M, roughly $0.97 a share. The 10-K warns that staked ETH "is subject to lock-up and unbonding periods during which it cannot be withdrawn or sold," and 5,227,037 ETH were staked on November 15, so selling ether to fund buybacks is slower than it looks. The fee stack is long-dated: the Ethereum Tower consulting agreement charges 1.00%, 0.50% and 0.25% of assets in tiers ($55.0M this year), runs ten years and owes 85% of the remaining fees if BitMine ends it without cause; the filing's note calls Ethereum Tower "the third-party service provider" and does not say who owns it. From September 4, American Validator LLC, which the 10-K calls "an affiliate of Ethereum Tower LLC," takes 1.5% of native staking rewards for ten years. Authorized common stock rose from 500 million to 50 billion shares in January, the preferred ranks ahead of common with a cash put on a change of control, and internal-control weaknesses remain open.
The Bottom Line
BitMine fixed its overhead this year and then gave the gain back on the side: from March on, revenue covered the cost of running the company, while the put program lost $143.7M across the year, more than the $121.4M the staking produced. What it has not fixed is the currency mismatch. Its income arrives as ether it keeps; its dividends, overhead and buybacks are paid in dollars from a cash pile that fell about $25M a week in early November, and the stock sale that refilled it now sells ether at roughly 70 cents on the dollar. Watch the September to December transition report for three things: whether the option program restarted, whether the buyback is funded by selling ether (Strategy's rule below 1 times, per a November 2 sec-api.io review of its 10-Q, is to sell coin rather than stock), and the share count against 626,451,021. Nobody wrote these figures by hand: every number about BitMine's own business came out of its SEC filings (the 10-K, its 10-Qs and its 8-Ks), with the share price and peer figures attributed where they appear, and the same reading works on any of the thousands of companies that file XBRL with the SEC, or on a private company that reports the same way.
Every figure about BitMine Immersion Technologies comes from its SEC filings, read through the RoboSystems SEC Shared Repository and verified against the filing with xbrlkit. Share price, analyst targets and peer multiples are attributed and dated where they appear.
Ask the filing yourself
Every figure in this report came from Bitmine Immersion Technologies, Inc.'s SEC filings in the RoboSystems SEC Shared Repository. Connect it to Claude or ChatGPT and ask your own questions, starting with this one:
How much did BitMine Immersion lose on derivative contracts in fiscal 2026, and how does that compare with its staking and validation revenue?