The biggest corporate holder of Bitcoin now sells stock and coin to pay its own shareholders

For six years, Michael Saylor's Strategy was a one-way valve: raise money from equity and bond investors, convert every dollar into Bitcoin, never sell. That valve has reversed. In the week of August 10 through August 16, the company sold 3.46 million MSTR shares for $333.7 million under its at-the-market program and bought no Bitcoin at all, splitting the proceeds between preferred dividends, a buyback of its STRC preferred stock, and its dollar reserve (Decrypt, Aug 18). It had done the same thing in July, selling roughly 4.8 million shares for about $466.7 million (Stocktwits, Aug 2026). The largest corporate holder of Bitcoin in the world is now raising cash from its own common shareholders to make interest payments to its preferred ones.

The trigger is arithmetic. Strategy holds 840,447 Bitcoin acquired for $63.36 billion, an average price of $75,385 per coin (Wireopedia, Aug 18). Bitcoin trades near $63,500 after US spot ETFs shed $389.7 million in the week of August 10 to 14, their worst weekly outflow in six weeks (247wallst.com, Aug 17). The treasury is underwater by roughly $10 billion. The company's stock has fallen about 39 percent this year while trailing the very ETF it was built to beat (24/7 Wall St., Aug 17).

The slow pressure underneath is the collapse of the premium that made the whole machine work. Strategy's market value once stood at three to four times the value of its coins; that ratio sat around 2.5x in December 2024 and has slid toward parity, with live trackers reading it near 1.03x (CoinMarketCap Academy, 2026; mnav.com, Aug 21). While that ratio traded far above one, the company could issue new shares at prices that added more Bitcoin per existing share — accretion by premium. At or below one, issuing shares dilutes owners without buying anything new. The flywheel did not seize. It started turning backward.

So Strategy pivoted from growth to maintenance. On July 6 it disclosed in an 8-K filing that it sold 3,588 Bitcoin for $216 million with an explicit stated purpose: paying dividends on its preferred shares (CleanSky, citing 8-K filing, Jul 6). That followed two smaller disposals earlier in the year, including 1,638 coins in its third sale of 2026, again funding preferred dividends and STRC buybacks rather than new purchases (ETHNews, Aug 2026). JPMorgan warned in early July that the new sales policy creates two-way risk for crypto markets, as STRC fell 25 percent below par in late June (TechTimes, Jul 3). Management responded by building a buffer: a USD reserve of $2.55 billion, which the company says covers 17.4 months of preferred dividends (Sandmark, Aug 2026).

Name the actors and what each wants. Saylor wants his borrowed-money thesis intact and his preferred stack paid, because default would end the story permanently. Common shareholders want the premium back and are being diluted to fund the people ahead of them. Preferred holders in STRK, STRF, STRD and STRC want their 8-to-10 percent yields, which is all they ever signed up for (Strategy.com investor materials, 2026). Bitcoin miners, squeezed by the same flat price, are selling their own coins into a market already absorbing ETF redemptions — Wintermute called it a supply pincer blocking any breakout (TechTimes, Aug 19). And passive ETF holders, who own the same asset without any dividend obligation, are simply leaving.

When you bolt a fixed income promise onto a volatile asset, the promise runs the company, not the asset.

The history that fits is the American investment trust boom of the late 1920s. Trusts like Goldman Sachs Trading Corporation traded at fat premiums to their portfolios, so promoters kept issuing new shares and stacking them into related entities, each issuance looking like proof of demand until the premiums broke and the entire structure unwound faster than the assets inside it ever fell.

What is different this time matters. Strategy's liabilities are perpetual preferred with no maturity date, not short-term paper, and the company controls the pace of its own selling. The counterexample argues the other way too — closed-end funds have traded at discounts for decades without cascading defaults, because none of them promised a fixed yield they could only meet by liquidating the portfolio. Strategy does exactly that.

Walk the consequences. First order: every dividend week now puts either newly printed MSTR shares or actual Bitcoin onto the market, mechanical supply regardless of price. Second order: the mNAV discount makes future capital raising harder just as it becomes more necessary, tightening the squeeze each quarter. Third order: if Bitcoin keeps drifting down, the $2.55 billion reserve starts drawing down, and the choice arrives in public — cut the preferred payout, sell coins at scale into a falling market, or both. The people who pay are common shareholders, whose claims get thinner with every ATM filing, and retail buyers of the preferred, whom JPMorgan flagged for deepening losses. The people who profit, for now, are the preferred yield collectors collecting on schedule, and the traders shorting the spread between the story and the balance sheet.

The observable sequence if this read is right: the weekly 8-K filings keep showing zero Bitcoin purchases stretching past the current eight-week drought (TechTimes, Aug 19), further coin sales appear whenever Bitcoin tests $60,000, and the USD reserve balance starts shrinking in the disclosures rather than growing. What breaks the read: Bitcoin reclaiming $75,000, Strategy's average cost, which would restore the premium, restart the buying machine, and turn every one of these maintenance sales into an embarrassing footnote. The whole argument lives or dies within about twenty percent of the current price.

The deeper point is that Bitcoin's corporate adoption story has quietly changed species. What was pitched as a treasury policy — a sovereign store of value held patiently forever — now behaves like a yield instrument with a collateral problem: obligations in cash at the top of the stack, a volatile asset below them, and a shrinking premium to hide the gap. Strategy set out to prove a corporation could hold Bitcoin better than a fund. It is ending the year proving something else: that when you bolt a fixed income promise onto a volatile asset, the promise runs the company, not the asset. The treasury became the debtor, and the debt learned to sell the treasure.