The largest corporate bitcoin holder now sits $9 billion underwater on 843,775 coins, has sold bitcoin for the first time in four years, and is funding preferred dividends from the asset it promised never to sell.
Summary
- Strategy reported an $8.22 billion net loss for Q2 2026, driven almost entirely by an $8.32 billion unrealized markdown on its bitcoin holdings under fair-value accounting, swinging from a $10.02 billion profit in Q2 2025.
- The company holds 843,775 bitcoin purchased at an average of $75,476 per coin, now worth roughly $54.8 billion against a $63.7 billion acquisition cost, a gap of approximately $9 billion.
- Strategy sold bitcoin for the first time in four years, disposing of 3,588 coins for $218.4 million to fund preferred stock dividends, and has authorized a program allowing up to $1.25 billion in future sales.
- The capital structure has shifted toward preferred equity, with $14.4 billion in preferred stock outstanding, annual dividend obligations approaching $1.2 billion, and cash reserves of $3.75 billion covering roughly 2.1 years of payments.
- MSTR shares declined 0.67% in after-hours trading, a muted reaction that reflects how thoroughly the market has internalized Strategy as a leveraged bitcoin proxy rather than a software company.
The Q2 numbers in context
Strategy’s Q2 2026 earnings report arrived on July 30 with an $8.22 billion net loss, a $24.45 loss per diluted share, and the kind of headline that writes itself. The number missed analyst estimates of negative $7.52 per share by a margin wide enough to qualify as a different conversation.
Set it against Q2 2025 and the swing is $18.24 billion in a single year: from $10.02 billion in net income to $8.22 billion in net loss. The underlying software business generated $122.39 million in revenue, roughly in line with the $122.91 million estimate and up 6.9% year over year. Subscription revenue grew 54%. Gross margin held at 66.6%.
None of that mattered. The software business is not why anyone owns this stock, and it has not been for years. The $8.32 billion unrealized markdown on bitcoin holdings is what produced the operating loss, and the operating loss is what produced the headline. Everything else is a rounding error on a balance sheet dominated by 843,775 coins.
The accounting rule that swings billions
The loss is real in an accounting sense and meaningless in an operational one, and understanding why requires understanding a single rule change.
In 2025, Strategy adopted ASU 2023-08, the Financial Accounting Standards Board’s fair-value standard for digital assets. Under the previous impairment model, companies marked bitcoin down when prices fell but could not mark it back up when prices recovered. The new standard requires marking to market at the end of every quarter and running the change, up or down, straight through net income.
When bitcoin rises, Strategy books a gain. When bitcoin falls, Strategy books a loss. No coins need to change hands. The $8.32 billion markdown in Q2 2026 reflects the decline in bitcoin’s price during the quarter, from roughly $86,000 at the end of Q1 to $64,915 at the end of Q2. The $10.02 billion profit in Q2 2025 reflected a price increase over that quarter.
The accounting treatment turns Strategy’s income statement into a bitcoin price chart with a six-digit multiplier. This is not a criticism of the standard; fair-value accounting is what the industry asked for, and it replaced a rule that was genuinely worse. Under the old impairment model, Strategy once carried its bitcoin at below $16,000 per coin on its balance sheet while the market price sat above $94,000. The new rule fixes that distortion. But it introduces a different one: every quarterly earnings report is now dominated by a number that tracks bitcoin’s spot price, and every headline leads with a figure that tells you nothing about whether the company can meet its obligations.
The swing between quarters illustrates how extreme this effect can be. In Q2 2025, bitcoin rose and Strategy booked $10.02 billion in net income, the largest quarterly profit in the company’s history. One year later, bitcoin fell and Strategy booked an $8.22 billion loss, the largest quarterly loss. The software business, the thing the company actually operates, generated roughly the same revenue in both quarters. The P&L swung $18 billion on the movement of an asset that was neither bought nor sold during the period.
For investors who understand the accounting, the earnings figure is noise. For headline readers, it is the story. And for analysts who must issue estimates, it requires predicting bitcoin’s end-of-quarter price, which is another way of saying it requires predicting the unpredictable. The $7.52 consensus estimate for the loss per share was off by more than three times, not because the analysts were wrong about Strategy’s business, but because they were wrong about where bitcoin would close on June 30.
843,775 coins and a $9 billion gap
The holdings are the thesis and the risk in a single number. Strategy holds 843,775 bitcoin purchased at an average price of $75,476 per coin, for a total acquisition cost of approximately $63.69 billion. At the reporting date price near $64,915, the portfolio was worth roughly $54.8 billion.
The gap is approximately $9 billion. Strategy is underwater on its aggregate position.
Year to date, the company acquired 29,997 additional bitcoin, growing holdings by 25% in 2026 and 11% quarter over quarter. The BTC Yield metric, which measures the growth in bitcoin per assumed diluted share, stood at 4.5% for the first half. Management frames this as the core performance indicator: not the price of bitcoin, but the rate at which the company accumulates more of it per share outstanding.
The framing is self-serving, but it is not without logic. If bitcoin’s price eventually exceeds the cost basis, the accumulation during the drawdown period represents buying at a discount. If it does not, the accumulation represents compounding a loss. The metric assumes an outcome and measures progress toward it. No traditional financial metric works this way, which is either the point or the problem, depending on your priors about bitcoin’s long-term trajectory.
The doctrine that broke
For four years, one rule anchored the most influential trade in crypto: Strategy buys bitcoin and never sells it. Michael Saylor said it in earnings calls, in interviews, in tweets that became doctrine. The promise was the spine of the thesis, the thing that made MSTR a leveraged bitcoin proxy rather than a fund that might trade around its position.
In late May, Strategy sold 32 bitcoin for $2.5 million. In late June and early July, it sold 3,588 bitcoin for $218.4 million. The proceeds funded preferred stock dividend payments. On June 29, the board formalized the shift with the Digital Credit Capital Framework, authorizing up to $1.25 billion in bitcoin sales to fund dividends, reserve maintenance, and debt service.
The never-sell era is over. Saylor has reframed his advice, saying “never sell your bitcoin” was directed at individual holders, not a corporate treasury commitment. Whether the distinction holds depends on whether you think the people who bought MSTR at $400 understood it that way.
The sales are small relative to the position. The 3,588 coins sold represent roughly 0.4% of holdings. But the doctrine was not about the size of the sales. It was about the certainty that there would be none. Once that certainty breaks, every future quarterly report invites the question: how much did they sell this time?
The Digital Credit Capital Framework formalized on June 29 makes the conditions explicit. The framework authorizes bitcoin sales for three purposes: funding preferred stock dividends, maintaining the USD reserve at target levels, and servicing debt obligations. It also established a $1 billion buyback program for common stock and a $1 billion buyback program for STRC preferred shares. The architecture is designed to give management flexibility in both directions, buying bitcoin when conditions are favorable and selling when obligations require it.
Saylor framed the shift on the Q1 earnings call, saying Strategy would “probably sell some bitcoin to pay a dividend just to inoculate the market and send the message that we did it.” The word “inoculate” is revealing. It treats the sale as a vaccine against future panic, a controlled exposure to the idea that Strategy can sell, so that when it sells again, the reaction is smaller. Whether the inoculation worked is testable: the stock declined less than 1% on earnings day, suggesting the market has absorbed the new regime.
The question that remains unanswered is scale. Selling 3,588 coins to cover a quarterly dividend is manageable. Selling bitcoin to cover a $1.5 billion annual dividend load, if issuance continues at the current pace, is a different proposition. The framework permits it. The thesis requires it not to happen.
The capital structure underneath
Strategy has built the most complex capital structure in crypto, and possibly the most unusual one on any major exchange.
On the debt side, $6.71 billion in convertible notes remain outstanding, down 18% after the company repurchased $1.5 billion at an 8% discount. On the equity side, $14.4 billion in preferred stock is outstanding, spread across multiple series. The company raised $8.41 billion in Q2 alone through at-the-market stock offerings: $2.95 billion from common shares and $5.47 billion from STRC preferred stock. Year-to-date capital raised: $17.06 billion.
The preferred stock is where the pressure lives. STRC, listed on Binance as the company expanded its funding channels, trades near par and pays an 11.5% annual dividend. Strategy calls it “Digital Credit” and frames it as a new asset class. In practice, it is a preferred share that funds bitcoin purchases and requires cash dividends that bitcoin appreciation alone cannot pay.
Preferred dividends paid to date total $1.06 billion. The annual obligation is projected to rise from $217 million in 2025 to $904 million in 2026. As the company issues more STRC to buy more bitcoin, the dividend load grows. This is the flywheel running in reverse: the mechanism that accelerated accumulation during the bull market now accelerates cash outflows during the bear market.
Cash reserves stand at $3.75 billion, which CFO Andrew Kang says covers existing dividend and interest obligations for “more than 2.1 years.” The target range is two to three years of coverage, a window the company has publicly committed to maintaining. That is adequate today. Whether it remains adequate depends on how much more STRC the company issues and what bitcoin does over those two years.
The company also authorized a $1 billion share repurchase program for common stock. No buybacks have been executed to date. The authorization exists as optionality, not as a signal of intent: buying back stock while simultaneously issuing new stock would be contradictory, and Strategy is still firmly in issuance mode.
The convertible debt reduction deserves separate attention. By repurchasing $1.5 billion of convertible notes at an 8% discount, Strategy reduced its fixed-income obligations while taking advantage of the notes trading below par. This shifts the capital structure from debt (with maturity dates and conversion triggers) toward preferred equity (with no maturity but perpetual dividend obligations). The trade-off is clear: less risk of a forced conversion event, more risk of a perpetual cash drain. Whether that exchange favors shareholders depends entirely on how long bitcoin stays below cost basis.
Why the stock did not move
MSTR shares declined 0.67% to $97.09 in after-hours trading. For a company reporting an $8.22 billion loss against a $2.15 billion estimate, a sub-1% decline is remarkably composed.
The reason is straightforward: MSTR does not trade on earnings. It trades on bitcoin, and bitcoin’s price was already known. The quarterly loss was baked into the share price the moment bitcoin fell below Strategy’s cost basis. The earnings report confirmed what the market had been pricing for months.
This is what the stock price prediction analysis identified as the core dynamic: MSTR is a bitcoin derivative with a management fee attached. The management fee is the dilution from continuous stock issuance and the dividend obligations on preferred shares. As long as bitcoin’s expected return exceeds that fee, the stock has a thesis. When it does not, the stock trades below the net asset value of the bitcoin it holds.
At recent prices, Strategy’s market capitalization has traded at or below the value of its bitcoin holdings for the first time. The premium that powered the accumulation flywheel, allowing the company to issue stock worth more than the bitcoin it could buy with the proceeds, has compressed to zero or turned negative. Without a premium, the flywheel does not work.
Software subscription revenue growing 54% year over year is a footnote in this context, but it matters for one reason: it provides roughly $500 million in annualized revenue that partially offsets the cash costs of the capital structure. Strategy is not purely a holding company. It has a business that generates cash, even if that business is now roughly 2% of the enterprise value conversation.
The VanEck analysis published before earnings noted that credit risk has fallen as preferred equity value now surpasses convertible debt, removing the forced-sale scenario that would occur if convertible notes matured without refinancing. That structural improvement is real, even if the headline number obscured it. The market, which lives in the details rather than the headlines, appears to have noticed.
The arithmetic that matters
The sustainability question reduces to a comparison between cash coming in and cash going out.
Cash in: $3.75 billion in reserves, plus software revenue of roughly $500 million annualized, plus the ability to raise more capital through stock issuance (though the premium compression limits how accretive this can be).
Cash out: preferred dividends projected at $904 million in 2026, plus convertible note interest, plus operating expenses. The bitcoin monetization program provides a release valve, but exercising it means selling the asset the company exists to accumulate.
Kang’s 2.1-year coverage ratio assumes no additional preferred issuance. But Strategy has been issuing STRC aggressively to fund bitcoin purchases: $5.47 billion in Q2 alone. Each new issuance adds to the dividend obligation. If the company raises another $5 billion in STRC in the second half, the annual dividend load could approach $1.5 billion, compressing the coverage ratio to roughly 18 months even before any new bitcoin is purchased.
The counter-argument deserves its strongest form. Bitcoin at $64,915 is below Strategy’s cost basis of $75,476, but it remains above the levels that would threaten the capital structure. Kang noted that cash reserves cover obligations for more than two years, and the company has multiple levers: it can slow accumulation, reduce preferred issuance, use the monetization program selectively, or wait for bitcoin to recover. Saylor’s pre-earnings statement that “governments can delay bitcoin adoption; they cannot stop bitcoin” reflects a conviction that the cost basis will eventually be exceeded.
If bitcoin returns to $100,000, Strategy’s unrealized loss becomes an unrealized gain of roughly $20 billion, the market implications reverse entirely, and the entire narrative flips. The 843,775 coins accumulated during the drawdown become the trade of the decade. The model depends on that outcome. The question is whether the capital structure can survive long enough to see it.
The 91% probability estimate that MSTR enters the S&P 500 adds a structural catalyst. Index inclusion would force passive buying from every S&P 500 tracker fund, providing a floor of demand independent of bitcoin’s price. Whether that probability holds after the earnings miss and the premium compression is a question the index committee will answer in coming months.
What to watch
- Bitcoin’s price relative to the $75,476 cost basis. Every dollar above that line turns the unrealized loss into an unrealized gain. Every dollar below it widens the gap and increases pressure on the capital structure.
- The STRC dividend coverage ratio. If Strategy continues issuing preferred stock while bitcoin stays below cost basis, the cash coverage window shrinks. Watch for Kang’s quarterly guidance on reserve duration.
- The bitcoin monetization program. The $1.25 billion authorization gives Strategy room to sell, but each sale erodes the accumulation thesis. The pace and size of future sales will signal how much pressure the dividend obligations are creating.
- S&P 500 index committee decisions. Inclusion would be the most significant structural catalyst for the stock since the bitcoin strategy began. Exclusion or delay would remove a source of expected demand.
- The premium or discount to net asset value. When MSTR trades above the value of its bitcoin, the flywheel works. When it trades below, the company cannot issue stock accretively. The mNAV ratio is the single best indicator of whether the model is functioning.
Frequently asked questions
How much did Strategy lose in Q2 2026?
Strategy reported an $8.22 billion net loss for Q2 2026, driven by an $8.32 billion unrealized markdown on its bitcoin holdings under fair-value accounting. The loss per diluted share was $24.45, against analyst estimates of negative $7.52. The underlying software business generated $122.39 million in revenue, roughly in line with estimates.
Why did Strategy report such a large loss?
The loss is almost entirely a result of ASU 2023-08, the fair-value accounting standard Strategy adopted in 2025. The rule requires marking bitcoin to market price at the end of each quarter. Bitcoin fell from roughly $86,000 to $64,915 during Q2, producing the $8.32 billion unrealized markdown. No significant amount of bitcoin was sold to produce the loss.
How much bitcoin does Strategy hold?
Strategy holds 843,775 bitcoin purchased at an average price of $75,476 per coin, for a total acquisition cost of approximately $63.69 billion. At the Q2 reporting date price near $64,915, the holdings were worth roughly $54.8 billion, approximately $9 billion below cost.
Did Strategy sell bitcoin?
Yes. Strategy sold 3,588 bitcoin for $218.4 million in late June and early July 2026 to fund preferred stock dividend payments. This was the company’s first significant bitcoin sale in four years. The board has also authorized a program allowing up to $1.25 billion in future bitcoin sales.
What is STRC?
STRC is Strategy’s perpetual preferred stock, branded as “Digital Credit.” It trades near its $100 par value and pays an 11.5% annual dividend. Strategy has issued $14.4 billion in preferred equity, with STRC as the primary vehicle. The dividend obligations on preferred stock are the main reason the company has begun selling bitcoin.
Why did the stock barely move after the loss?
MSTR shares declined only 0.67% after hours because the market already knew bitcoin’s price. The quarterly loss was a function of bitcoin falling below Strategy’s cost basis, which had been reflected in the stock price for months. MSTR trades as a leveraged bitcoin proxy, and the earnings report contained no new information about bitcoin’s trajectory.
What is BTC Yield?
BTC Yield is Strategy’s proprietary metric measuring the growth in bitcoin holdings per assumed diluted share. It stood at 4.5% for the first half of 2026, reflecting the addition of 29,997 bitcoin year to date. Management frames this as the core performance indicator, though critics note it assumes bitcoin’s price will eventually exceed the cost basis.
Is Strategy’s model sustainable?
The model depends on bitcoin eventually exceeding Strategy’s $75,476 average cost. Cash reserves of $3.75 billion cover dividend and interest obligations for approximately 2.1 years. If bitcoin recovers, the accumulated position becomes enormously profitable. If it does not, and if the company continues issuing preferred stock, the dividend obligations could exhaust cash reserves and force larger bitcoin sales. The company has levers to manage this, including slowing accumulation and using the monetization program, but the capital structure is more leveraged to bitcoin’s price than at any point in the company’s history.
Disclaimer: This article is for informational purposes only and does not constitute financial, investment, or trading advice. Strategy’s stock and bitcoin are volatile assets with significant risk of loss. The information presented reflects the state of events as of July 31, 2026, and may change as market conditions develop. Readers should consult qualified professionals before making investment decisions.
15086#Strategy #posts #billion #loss #bitcoin #falls #cost #basis1785502542
