The number is 63,957. That is the average execution price, in dollars, at which Strategy sold 1,638 bitcoin in August 2025. The sale produced $104.7 million in proceeds. The company's average acquisition cost for its bitcoin treasury is approximately $75,419 per coin. On this tranche, the company realized a loss of roughly $18.8 million.
It did not sell to reduce risk. It did not rotate into another asset. It sold to service a dividend. Approximately $52.4 million went to STRC preferred shareholders as a coupon payment. Approximately $52.3 million went to retire 912,143 STRC shares. In the same window, the company issued 3,011,361 new MSTR common shares, raising approximately $290.6 million in net proceeds.
The accounting is simple. The consequences are not.
I have spent eleven years auditing this industry's balance sheets. I spent 72 hours tracing Anchor Protocol's yield mechanics in 2022 and published a 40-page report proving that yield was debt, not revenue. I traced $4.5 billion of FTX assets across five chains in late 2022. I follow one professional rule: when the narrative and the ledger diverge, the ledger wins.
The August 8-K is the ledger. It states that the world's largest public bitcoin holder sold bitcoin at a loss to service a preferred-stock coupon. That is not a treasury strategy. That is a liability schedule.
Context: The Structure Under Pressure
The broader tape matters here. Bitcoin has been range-bound through the consolidation phase that defined mid-2025. In a sideways market, cash-flow obligations expose themselves. During a bull phase, no one audits the coupon. During a flat phase, the coupon audits everyone.
Strategy, formerly MicroStrategy, remains the largest publicly traded corporate holder of bitcoin. After the August disposal, the treasury holds 842,138 BTC. That is 4.01% of the 21 million supply ceiling. It is roughly 2.4 times the bitcoin held by BlackRock's IBIT ETF and orders of magnitude beyond any other public company.
The company's model since 2020 was deceptively simple: issue equity or convertible debt, buy bitcoin, hold. The MSTR premium — the gap between the company's market capitalization and the market value of its bitcoin holdings — was sustained by the expectation of continuous accumulation. The market priced Strategy as a permanent bid.
In early 2025, the company introduced a second instrument. STRC is a perpetual preferred stock with a fixed annual dividend of 12%. Disclosures specify a payment of $0.50 per share per semi-annual period. The market now prices this instrument below par: STRC trades near $92 against a $100 face value. A preferred stock trading below its face value is a verdict on coupon sustainability. It is a discount applied to probability-weighted future cash flows.
Two further facts frame the event. First, the company has not purchased bitcoin for five consecutive weeks. That is the longest pause since the accumulation campaign began. Second, the board approved a capital framework in June 2025 permitting bitcoin sales of up to $1.25 billion. The stated plan is to raise that ceiling to $5 billion.
The August sale is not an isolated event. It is the first execution of that framework. The prior model: raise capital, buy bitcoin, appreciate, raise cheaper capital, repeat. The current model: sell bitcoin, generate cash, service preferred obligations, dilute common equity, repeat. The direction of causality has inverted.
This is a balance-sheet event, not a protocol event. There is no bytecode to audit. There is no smart contract. The terms of the preferred instrument are the code, and the code has a deterministic flaw.
Core: The Mechanisms, Dissected
I will trace the mechanics line by line.
Line-by-Line Reconciliation
The August transaction has four legs. Leg one: 1,638 BTC sold at approximately $63,957 per coin, gross proceeds $104.7 million. Leg two: $52.4 million disbursed as the STRC dividend. Leg three: $52.3 million applied to STRC repurchases, retiring 912,143 shares. Leg four: 3,011,361 MSTR shares issued, net proceeds approximately $290.6 million.
The split is nearly precise: 50.1% dividend, 49.9% buyback. That ratio is not accidental. The sale was sized to cover a pre-existing obligation schedule, not to express a view on bitcoin's price.
One data-integrity note: the disclosed repurchase figures do not fully reconcile. The stated buyback tranche is $52.3 million for 912,143 shares, implying approximately $57.30 per share, while other disclosures reference a total repurchase consideration of $81.2 million. The Form 8-K does not reconcile these figures. I flag this the same way I flag a failing test in a formal verification report: the components do not sum, therefore the review request goes out. The structural conclusion does not depend on the reconciliation. The discrepancy is still material enough that no audit partner should ignore it.
The Dividend Is a Liability, Not a Narrative
The market's core error is the belief that a 12% preferred dividend is an income stream. It is not. It is a contractual obligation with a defined payment schedule and a defined priority.
The arithmetic is unforgiving. If the aggregate outstanding par value of STRC is approximately $5.8 billion — consistent with the disclosed payout scales and the $4 billion reserve sizing — the annual dividend obligation approaches $700 million. That is $1.92 million per day, whether bitcoin rises, falls, or trades flat. The obligation matures every six months, forever.
Strategy's legacy software business does not generate free cash flow at that scale. The company's financial engine was never operations. It was balance-sheet appreciation. As long as bitcoin appreciated, the coupon was trivially serviceable: the reserve asset grew faster than the coupon accrued.
That condition no longer holds. Bitcoin trades below the company's average acquisition cost. In this regime, the dividend must be sourced from one of three places: sale of the reserve asset, issuance of common equity, or drawdown of the cash reserve. The August transactions used all three. The company is not managing a treasury. It is servicing a coupon.
The Perpetual Problem
The word "perpetual" deserves scrutiny. A perpetual preferred has no maturity date. The issuer is never contractually released from the coupon. The only exit paths are repurchase, conversion, or default.
This is the structural flaw in a non-bull market. A perpetual obligation is a short position on the issuer's future cash-generating ability. When the issuer's only material asset is a volatile commodity, the perpetual coupon becomes a short position on that commodity's volatility. Every drawdown forces liquidation. Every liquidation locks in realized losses. Every realized loss reduces the equity cushion. Every reduced cushion forces the next liquidation to be larger in relative terms.
I audited this exact mechanism in 2022. Anchor Protocol promised a fixed 20% yield on UST deposits. The yield was not derived from productive assets. It was derived from a reserve being depleted at a compound rate. The model worked until inflows stopped. Then the reserve math took over. The outcome was determined before the narrative accepted it.
The same shape is present here. I am not comparing Strategy to a Ponzi. Anchor was a fraudulent structure. Strategy is a legitimate public company executing disclosed transactions. The similarity is narrower and more instructive: fixed obligations against a volatile asset base produce deterministic failure modes when the asset price does not cooperate. The difference is that Anchor had no underlying assets. Strategy holds 842,138 bitcoin. That is a real buffer. It is also a buffer that a 12% coupon can consume, given sufficient time.
The Two-Axis Dilution
The per-share bitcoin metric is the ratio of bitcoin holdings to outstanding shares. The August transactions degrade this ratio on both axes simultaneously.
Axis one: holdings decline. The company sold 1,638 BTC. The numerator decreases. Axis two: share count increases. The company issued 3,011,361 new shares. The denominator increases.
The combined effect is a reduction in bitcoin per share that exceeds either action alone. The equity raise produced $290.6 million. At current prices, that capital could have purchased approximately 4,500 BTC. It was not spent on bitcoin. It was spent, in significant part, servicing the preferred structure. Capital that once would have been accretive to the per-share ratio is now neutral to dilutive.
The bulls will object that issuance at a premium to net asset value is accretive regardless of deployment. That is correct only if the premium persists. A premium is not a constant. It is a variable derived from the market's confidence in future accumulation. The company has now demonstrated that future accumulation is conditional. The premium will re-price accordingly.
The Runway Calculation
The company holds a $4 billion USD Reserve. This is the disclosed liquidity buffer for dividend and operating obligations.
The math is a straightforward runway calculation. If the annual dividend obligation is approximately $700 million, the reserve covers roughly 5.7 quarters of dividends alone. That assumes zero further buybacks, zero bitcoin purchases, and zero additional calls on the reserve. Add the observed buyback cadence — approximately $52 million quarterly — and the runway shortens to roughly 4.6 quarters.
This is not an insolvency forecast. It is a term structure. The reserve is a five-quarter asset, not a permanent one. Replenishment is available through equity issuance, as August demonstrated. But the replenishment cost is measurable: dilution. The true cost of the coupon is not the $52.4 million quarterly payment. It is the sum of foregone bitcoin purchases, realized sale losses, and offsetting issuance dilution.
The Signal Is Not the Volume
Assess the market impact honestly. The 1,638 BTC sale is approximately 0.19% of the company's holdings and between 0.2% and 0.3% of average daily bitcoin volume. The trade itself is a rounding error.
The signal is the pause. Five weeks without a purchase. An authorized sale framework of $1.25 billion. A stated plan to raise the ceiling to $5 billion. The market priced Strategy as a permanent demand bid. That assumption is falsified. The bid has become conditional.
Quantify what the pause means. In the 2024-2025 cycle, Strategy was the single largest public buyer, with monthly acquisitions in the tens of thousands of coins. That recurring demand was a structural bid. Remove it, and the order book must absorb that absence. In 2023, I analyzed the Azuki ecosystem's spin-off collections and found that 60% of trading volume came from a single entity operating 15 wallets. The volume looked like demand. It was not. Distinguishing apparent demand from actual demand is the core discipline of market integrity. A large buyer who stops buying is not a small seller who sells. The first changes expectations. The second changes only the tape.
The market should read the pause as the primary data point. The sale is confirmation bias. The pause is the news.
The Negative Feedback Loop
STRC trades at approximately $92, an 8% discount to face value. The discount persists because the market impounds coupon-solvency risk. The company's response is to buy the instrument back. The buyback supports the price. The buyback is funded by the asset whose weakness created the discount.
The loop propagates. Discount persists. Buybacks accelerate. Cash is consumed. Cash is sourced from asset sales and equity issuance. Both reduce the per-share claim on the treasury. Common equity weakens. The next equity raise becomes more expensive. The company relies more heavily on asset sales. The market observes the selling and widens the discount.
Each step is individually rational. The aggregate trajectory is not sustainable. This is the reflexive structure I documented during the Luna collapse: the stabilization mechanism and the destabilizing force are the same transaction.
Governance: The Priority Inversion
Michael Saylor exercises outsized control over Strategy's direction. The company is, operationally, a vehicle for his bitcoin thesis. Concentration is a feature in the bull case: a single conviction holder willing to buy through drawdowns. It becomes a liability when decision quality is questioned. Selling at $63,957 against a cost basis of $75,419 is poor execution timing by any measure. It may be forced by the coupon schedule. It is still a realized loss that a more flexible capital structure would have avoided.
The deeper issue is the priority conflict. The company sold its core asset — the source of common equity value — to service a preferred obligation. Common shareholders bear the cost. Preferred shareholders receive the benefit. That is a textbook priority inversion: residual equity liquidated to support a fixed-income instrument trading below par.
Shareholder litigation is a plausible tail risk. The theory writes itself: the board authorized a 12% dividend structure, then authorized the sale of the company's principal asset to fund it, while diluting the common class. Whether that theory prevails is a question for courts. The exposure exists.
The reporting optics compound the problem. The company reported a second-quarter net loss of $8.22 billion, including an $8.32 billion unrealized write-down on bitcoin. Fair-value accounting does not change the cash position. It does change the equity cushion, the proximity of debt covenants, and the market's valuation framework.
The Regulatory Frame
STRC satisfies the four prongs of the Howey test: investment of money, common enterprise, expectation of profit, and efforts of others. A 12% fixed dividend is an explicit profit expectation. The instrument trades under SEC registration on NASDAQ, so the securities classification question is largely settled. The open question is disclosure adequacy.
If the company knew, or should have known, that the coupon was unsustainable without selling its principal asset, the sustainability of the coupon becomes a material disclosure item. The August 8-K discloses the transaction. It does not disclose the forward analysis: how much more bitcoin will be sold, at what price the dividend becomes unserviceable, or where the equity issuance schedule ends. Material omissions are harder to litigate than misstatements. They are still litigated.
The SEC's position on accounting treatment matters as well. Under fair-value guidance, Strategy's quarterly profit and loss statement will swing with bitcoin's dollar price. That mechanically injects volatility into reported earnings, which feeds multiple expansion or contraction in the common equity. An impairment line is not a cash event. It is a perception event with compounding consequences.
The Competitive Set
Place Strategy against the other large holders. BlackRock IBIT holds approximately 350,000 BTC in an ETF structure with low fees and direct redemption mechanics. Galaxy Digital holds roughly 50,000 BTC inside a diversified regulated financial services business. Tesla holds approximately 9,720 BTC, a relic of an abandoned experiment. Strategy's 842,138 BTC dwarfs all of them. But the encumbrance differs sharply. IBIT's coins belong to ETF shareholders who can redeem them at net asset value. Tesla's coins sit unencumbered and ignored. Strategy's coins are contractually linked to a perpetual 12% coupon. The same asset, three different risk profiles. The encumbrance is the differentiator.
This matters for the ecosystem. Strategy was a demand-side anchor for miners seeking over-the-counter exit liquidity. When the anchor becomes a conditional seller, miners lose a structural counterparty precisely at the post-halving moment when block rewards have shrunk. The chain of transmission runs from the coupon ledger to the mining revenue schedule. Indirect. Real.
Contrarian: What the Bulls Got Right
I have presented the bear case. Now I will argue against it, because the bulls are not wrong. They are conditionally right.
First: the sale is immaterial. 1,638 BTC is 0.19% of the treasury. Even the full worst-case liquidation — the $5 billion ceiling at current prices, approximately 78,000 BTC — leaves the company with more than 760,000 BTC. The core thesis, that Strategy is the largest public accumulator of bitcoin, survives the extreme scenario.
Second: the buyback is rational. Retiring a 12% perpetual obligation at a discount to par is statistically positive carry. The company is buying a dollar of future coupon obligations for approximately $0.92. If the underlying asset survives, that is a high-conviction use of capital. The buyback is not the problem. The dividend is the problem. They are separable instruments.
Third: the process was disciplined. The board authorized $1.25 billion. Management executed within the authorization. Disclosures were timely, filed on Form 8-K. Process integrity is rare and valuable. A transparent, framework-constrained sale is categorically different from a silent liquidation.
Fourth: issuance at a premium to net asset value can be accretive. If MSTR trades above bitcoin NAV, issuing shares at a premium to retire a 12% liability improves the common equity's claim on the treasury. The market will not reward this in one quarter. It will reward it across the full coupon-reduction cycle.
Fifth: the reserve is real. The $4 billion USD Reserve is a disclosed balance-sheet fact. It buys time. Time is the one input that cannot be manufactured by tokenomics or narrative. The company has purchased roughly five quarters of optionality.
There is also an information-efficiency argument. STRC at $92 is the market performing its job: discounting a stressed coupon. MSTR at its prevailing level is the market pricing a smaller accumulation premium. The system is functioning. The problem is not the market's perception. The problem is the instrument's design.
The bulls are correct that this is not fraud, not a liquidity crisis, and not a protocol failure. It is a stressed capital structure responding to a flat market.
My objection is narrower. The bulls treat the August sale as an isolated event. It is not. It is the first execution of a framework designed to permit more selling. The framework is the story. The August 8-K is the first page. A rational buyback does not address the irrational coupon that necessitated it. The company is solving an interest-rate problem by selling the principal.
Takeaway: The Three Variables
The test is now defined. Over the next two quarters, I will watch three variables.
First: the weekly 8-K disclosures. A resumption of purchases regenerates the thesis. A continuation of sales confirms the conversion.
Second: the STRC bid. A sustained decline below $90 is the market's verdict that the coupon is not credible at the disclosed rate. The company's response — accelerated buybacks, or a path to a lower coupon — will determine the instrument's future.
Third: the $5 billion ceiling. A board vote to raise the sale limit is the definitive confirmation that the treasury model has shifted from accumulation to servicing.
Bitcoin's long-term value is not the question. The question is whether a public company can carry a 12% perpetual dividend through a non-bull market. The mathematics say: only by selling the asset that justifies the company's existence. That is the difference between a treasury and a treadmill.
Trust is a variable. Proof is a constant. The August 8-K is proof. Adjust your models accordingly.
