Strategy did not sell 1,638 Bitcoin because it ran out of money. That misconception is worth dismantling first, because it drives almost every take on this story. The company holds a $4.0 bil
Strategy did not sell 1,638 Bitcoin because it ran out of money. That misconception is worth dismantling first, because it drives almost every take on this story. The company holds a $4.0 billion US dollar reserve, and Polymarket prices the odds of a 2026 margin call at 4.25%. This was not distress. It was arithmetic. Strategy's common stock now trades at roughly 0.68 times the value of the Bitcoin behind it — a 30% discount, down from about 3.4x in November 2024, according to Protos. Once that multiple falls below 1, issuing shares to raise cash gives away more Bitcoin claim than the cash is worth. Selling Bitcoin becomes the cheaper source of funds. That inversion — not conviction, and not capitulation — now governs when Strategy sells. And the company has written the rule down.
Here is the number that makes the cadence predictable. Strategy's own 10-Q, filed 3 August 2026, states that its expected annual preferred dividend and interest payments are approximately $1.76 billion. That is $4.82 million a day. At Bitcoin's 5 August price of $64,082 (CoinGecko, cross-checked against Coinbase at $64,095 and Kraken at $64,098), the bill converts to 27,465 BTC a year — 2,289 a month, or 528 a week. Strategy sold 1,638 BTC, which covers about 3.1 weeks of it. The sale was not a strategic reversal. It was an invoice, denominated in Bitcoin. Which means the real question was never whether Strategy sells again, but what cadence the invoice arrives on and what would stop it.
Bitcoin has closed below Strategy's $75,419 average cost basis on 144 of 218 trading days in 2026, and every single day since 27 May. Source: CoinGecko daily BTC/USD close, retrieved 5 August 2026. Most coverage stops at "Strategy sold Bitcoin to pay dividends," which is true and explains nothing. The mechanism is where the answer lives.
STRC — formally the Variable Rate Series A Perpetual Stretch Preferred Stock — carries a $100 stated amount and a dividend rate the board resets every 30 days. Perpetual means there is no maturity: the obligation never retires itself, and Strategy can only be rid of it by buying the shares back. The rate is the instrument's price-control valve. The prospectus is unusually candid about this, stating that the company's intention is "to adjust the monthly regular dividend rate per annum in such a manner as we believe will maintain STRC Stock's trading price at or close to its stated amount of $100 per share" — cutting the rate if it trades above $100, raising it if below.
Think of it as a mortgage where the interest rate rises every time your house appraises lower. When Bitcoin falls, the market demands more yield, so the cash cost of the structure rises precisely when the asset backing it is worth least. The rate has moved from 9.00% at launch to 12.00% effective July 2026. Contractually, increases are unconstrained; cuts are capped at 25 basis points plus any decline in SOFR. The valve turns up easily and down slowly.
Two features make the obligation non-negotiable. STRC dividends are cumulative: a missed payment accrues as arrears that must be cleared before a cent reaches common shareholders. And since 30 June 2026 they are paid semi-monthly — 24 payments of $0.50 rather than 12 of $1.00 — so cash leaves the building twice a month regardless of Bitcoin's price. STRC alone represents $10.49 billion of notional across 104.89 million shares, roughly $1.26 billion of the $1.76 billion bill. The full mechanics of STRC's dividend and par-value design are worth understanding before judging any of this.
Then, in June 2026, Strategy did something that changed the game and has been almost entirely missed. It partly de-linked the rate from the price. The 10-Q states that the company "will not necessarily increase the STRC dividend rate solely because STRC trades below its $100 per share stated amount," and that management does not intend to recommend a change from 12.00% "until STRC Stock demonstrates sustained trading at or near its $100 stated amount." Read that carefully: rather than keep ratcheting the coupon to defend the price — which would permanently enlarge the cash bill — Strategy chose to defend it by buying the stock back. Buybacks need cash. Cash, at a 30% equity discount, is cheapest to raise by selling Bitcoin. That single policy choice is the proximate cause of the 3 August sale.
Michael Tanguma, CEO of Onramp, framed the structural risk fairly: "A capital structure that survives volatility only by adding permanent obligations is a structure with a finite number of cycles in it." Preferred shares of this kind are not debt and cannot trigger a default — but they are not free equity either.
Strategy has stopped improvising. On 29 June 2026 the board adopted a Digital Credit Capital Framework with five pillars: a USD Reserve policy, a revised STRC dividend policy, a preferred repurchase programme, a common repurchase programme, and a BTC Monetization Program. A month later the company confirmed that new capital would no longer be allocated entirely to Bitcoin. For a company whose identity since 2020 was "buy and hold," that is the sentence of the year.
The framework contains the answer to "when," in writing. The 10-Q says the board authorised Bitcoin sales for three purposes: to generate up to $1.25 billion for the USD Reserve; to fund preferred dividends and interest "when management determines that doing so is more advantageous than issuing class A common stock"; and to fund repurchases of preferred or common stock. That middle clause is the trigger. It is not a price level and not a crisis threshold — it is a relative-cost test between two funding routes, and at a 30% equity discount that test resolves in Bitcoin's favour every time.
There is also a floor. The USD Reserve policy requires Strategy to hold cash equal to at least 12 months of expected preferred dividends and interest — about $1.76 billion — with any dip below that needing fresh board authorisation. The reserve stands at $4.0 billion, which is 27 months of coverage, leaving roughly $2.24 billion of discretionary headroom before the floor binds. Notably, none of the $1.25 billion reserve-building tranche had been used as of 24 July. Every sale so far has been for dividends and buybacks, not for building cash.
The capital flows that same week show the pivot plainly. Strategy sold 3,011,361 MSTR shares under its at-the-market programme, raising $290.6 million — and almost none of it bought Bitcoin. Instead $250 million lifted the dollar reserve, $28.9 million funded STRC repurchases and $11.7 million went to cash. The equity machine still runs at full speed; it simply no longer points at Bitcoin.
Michael Saylor's response was to separate the man from the company. "When I say 'Never Sell Your Bitcoin,' I speak as one saver to another. I have never sold mine," he wrote on 3 August. "Not one satoshi. Strategy is a public company, not my wallet. Since 2020, it has disclosed it may buy or sell $BTC to manage capital." Read uncharitably that is a retreat; read accurately it is a chairman noting that fiduciary duty to preferred holders is not personal ideology — a distinction the market spent five years failing to price.
Chief executive Phong Le was asked directly on CNBC's Power Lunch the same day whether a $75,000 average cost against a $63,000 spot price made him a forced seller. Le replied: "No, we're not a forced seller of bitcoin. The average cost is not that important to us... sometimes it makes sense to sell bitcoin to add to our U.S. dollar reserves. Sometimes it makes sense to pay dividends. And that's what we've been doing."
The argument for the other side deserves more than a dismissal, because it is quantifiable. Strategy repurchased 912,143 STRC shares for $81.2 million — roughly $89.02 per share against a $100 stated amount. That retires $91.2 million of notional at an 11% discount and permanently removes about $10.95 million a year of obligation. Every share retired shrinks next year's forced selling. If buybacks outrun the coupon, selling 1,638 BTC today buys back more than 1,638 BTC of future sales.
Joe Burnett, VP of Bitcoin Strategy at Strive, made the durability case on the earlier July sale, telling CoinDesk: "Strategy just sold ~1.5 months of dividend obligations in one week. At this pace and with 0% BTC appreciation, today's dividend obligation is funded until 2056... At ~3.4% annual BTC appreciation, today's dividend obligation can be funded indefinitely." His arithmetic checks out against the filings — $216 million against a $146.7 million monthly bill is 1.47 months. At 27,465 BTC a year, servicing the preferreds entirely from the treasury would take about 31 years to exhaust 842,138 coins.
Run the financing decision through the discount. If Strategy issues $1 billion of stock at an mNAV of 0.68, it receives $1 billion of cash — and hands new shareholders a claim on roughly $1.47 billion of Bitcoin. Issuing equity destroys about 47 cents of Bitcoin claim per dollar raised. Selling Bitcoin converts the asset at 100 cents. Below an mNAV of 1, selling the treasury is not the desperate option; it is the shareholder-friendly one. This is what every "Saylor capitulates" take gets backwards.
One nuance matters, and almost nobody reports it. The common stock trades at roughly 0.68x Bitcoin NAV, but the enterprise measure — including debt, cash and preferred — sits near 1.02–1.06x. The whole capital structure prices at about par. The 30% discount is not the market saying Strategy's Bitcoin is worth less; it is the market saying the preferred holders own that slice. Strategy now reports enterprise mNAV rather than the simple version, and redefined the metric outright on 23 July 2026 — its own filings warn that earlier figures are not comparable, so treat any cross-period mNAV chart with suspicion.
The market's verdict on the sale is the tell. Strategy disclosed before the open on 3 August. MSTR closed that session at $94.86, up 1.7% from $93.28, then rose a further 2.94% to $97.65 on 4 August — roughly 4.7% across two sessions, after announcing it had sold Bitcoin below cost. That is not how a tape reacts to distress.
Funding routeCost at mNAV 0.68Effect on Bitcoin per shareIssue common equity (ATM)$1bn raised gives away ~$1.47bn of Bitcoin claimDilutive — fallsIssue more preferredAdds a permanent 12% cash obligationNeutral now, worse laterSell BitcoinConverts at 100 cents; no new obligationFalls, but least of the threeBuy back STRC below parRetires notional at ~89c; needs cash from one of the aboveImproves future Bitcoin per shareThat gap is the metric most treasury-stock coverage ignores, and it drives every decision above. Strategy is not alone in facing it: Metaplanet has weighed emergency measures after its own mNAV sank. This is a model-wide stress test, not a Saylor problem.
The most useful skill in this story is not reading filings. It is reading chains — because the filing arrives days after the fact. Strategy executed the sale between 27 July and 2 August, and disclosed it on 3 August. For up to a week the transaction existed on the blockchain and nowhere in public documents. On-chain observers can close that gap; filing readers cannot. This is why retail has quietly reorganised around wallet-watching, and why a Reddit thread flagging an $18.91 million transfer to a fresh wallet called the direction before the 8-K did.
The method is straightforward. Strategy's treasury addresses are clustered and labelled by attribution platforms such as Arkham. The signal is not movement between the company's own cold-storage addresses but deposits into exchange or prime-broker addresses — Coinbase Prime above all. Then check the calendar. Every weekly Bitcoin and ATM update since May 2026 has been filed on a Monday, covering the prior Monday-to-Sunday period, which makes a mid-week exchange-bound transfer a candidate for confirmation the following Monday.
Now the discipline, because this is where retail gets hurt. A transfer is not a sale. A suspected Strategy-linked wallet moved another 1,030 BTC worth $66.14 million on 5 August, with no sale confirmed. Earlier this year a 411.48 BTC transfer to Coinbase Prime was widely read as an imminent liquidation; no sale followed. Prime brokers provide custody and financing as well as execution, so a deposit is equally consistent with posting collateral. The correct inference from a large transfer is "the odds of a Monday disclosure just rose," never "they sold."
No regulator is threatening Strategy's treasury. The pressure comes from two quieter directions, both more consequential than anything the SEC is doing.
The first is accounting. Under the fair-value standard adopted in 2025, digital assets are marked to market every reporting period. That rule produced the $8.22 billion second-quarter loss, of which $8.32 billion was digital-asset losses and — tellingly — under $1 million was realised. A company generating $122.4 million of growing software revenue reported one of the sector's largest quarterly losses because of a price quote. The accounting is doing real economic work here: headline losses depress the equity, which widens the discount, which makes Bitcoin sales the rational funding route.
The second is index governance, and it is badly under-covered. Index providers deciding whether digital-asset treasury companies belong in mainstream benchmarks are, functionally, writing regulation without a rulemaking process. Polymarket prices Strategy's removal from the MSCI index by 31 December at 56.5%, up seven points on the day. Deletion would force passive funds to sell MSTR mechanically, pushing the share price down, widening the discount, making equity issuance more punitive still — and making Bitcoin sales more attractive. An index committee would tighten the same screw the dividend already turns. JPMorgan has warned this policy could amplify crypto volatility; this is the channel through which it happens.
Yes, and the mechanism gives a cadence rather than a date. Three calls, with the reasoning attached.
1. Expect roughly 1,500–2,300 BTC sold per month while the discount persists. The $1.76 billion bill converts to 2,289 BTC a month at current prices, and the run rate scales inversely with Bitcoin: a fall to $50,000 lifts it to about 2,930 BTC a month. The next weekly 8-K is due Monday 10 August, covering 3–9 August — every weekly update since May has landed on a Monday. Note what this rate is not: at 27,465 BTC a year it would take roughly 31 years to exhaust the treasury. This is a slow bleed, not an unwind.
2. Treat the 56% headline with real caution — that market is thin. Polymarket's 4–10 August sell market sat at 56% on 5 August, having jumped 19 points in a day, against 10.5% for a purchase. But it had traded under $1,000 of total volume with roughly $283 of liquidity and a 48c–64c spread. Compare the margin-call market: $98,000 of volume, priced at 4.25%, stable. Read the 56% as loose sentiment from a handful of traders; read the margin-call price, which is deep, as the market's real judgement. Solvency is not the issue.
3. Sales stop when the discount closes, not when Bitcoin recovers. This is the disconfirmation test and it matters more than the price level. Bitcoin could rally to $80,000 and sales continue if the equity discount persists; equally, a re-rating of MSTR toward parity would restore accretive issuance and purchases could resume with Bitcoin still in the sixties. Watch the ratio, not the coin. The secondary trigger is STRC's own price: sustained trading near $100 would end the buyback programme that is currently consuming the cash, and Strategy has said explicitly it will not revisit the 12.00% rate until that happens.
The larger point is that Strategy has stopped being a leveraged Bitcoin proxy and become something closer to a closed-end fund with an expensive perpetual financing layer. That is a less thrilling instrument, but a legible one. The company still holds 842,138 BTC, still has $4.0 billion of cash against a $1.76 billion policy floor, and still faces almost no risk of forced liquidation. It will simply keep selling small, regular amounts of Bitcoin for as long as its shares are worth less than the Bitcoin behind them.
No. Strategy holds a $4.0 billion dollar reserve against $1.76 billion of annual preferred dividends and interest — 27 months of coverage against a 12-month policy floor — and Polymarket prices a 2026 margin call at 4.25%. CEO Phong Le said directly on CNBC on 3 August: "No, we're not a forced seller of bitcoin." The sales are a capital-allocation choice driven by a 30% equity discount, not a liquidity emergency.
mNAV compares a company's market value with the value of its digital-asset holdings. Above 1, issuing shares buys more Bitcoin per existing share — accretive. Below 1, issuing shares hands new investors more Bitcoin claim than the cash raised is worth. Strategy's common trades near 0.68x, so a $1 billion raise gives away about $1.47 billion of Bitcoin claim. That flips selling Bitcoin from a last resort into the rational option.
STRC is Strategy's Variable Rate Series A Perpetual Stretch Preferred Stock, with a $100 stated amount and a rate the board resets every 30 days. It launched at 9.00% in July 2025 and now pays 12.00%, semi-monthly and cumulatively, in cash. Since June 2026 Strategy has defended STRC's price by repurchasing shares rather than raising the rate further — and those buybacks need cash the equity market cannot cheaply provide.
Approximately 842,138 BTC after the 3 August 2026 sale, acquired for $63.51 billion at an average $75,419 per coin. At $64,082 that position is worth about $53.97 billion, an unrealised loss near $9.6 billion. Strategy has sold roughly 5,258 BTC in 2026 — about 0.6% of its holdings — and remains by a wide margin the largest corporate Bitcoin treasury in the world.
Partially. On-chain transfers appear days before the filing: the 3 August disclosure covered trades executed from 27 July. Deposits into exchange or prime-broker addresses such as Coinbase Prime raise the probability of a disclosure the following Monday, since every weekly 8-K since May 2026 has been filed on a Monday. But transfers are not sales — a 411.48 BTC move to Coinbase Prime earlier in 2026 was never followed by a confirmed sale.
He says no. "I have never sold mine. Not one satoshi. Strategy is a public company, not my wallet," Saylor wrote on 3 August 2026, drawing an explicit line between his personal holdings and the company's balance-sheet decisions. Strategy has disclosed since 2020 that it may buy or sell Bitcoin to manage capital, so corporate sales do not contradict his personal position.
This article is for information purposes only and does not constitute investment advice. Prices and prediction-market odds cited were retrieved on 5 August 2026 and move quickly.
