The unrealised loss reflects the gap between Strategy's $75,578 average cost and quarter-end bitcoin prices near $58,700. STRC preferred grew to $10.5 billion notional.
Strategy reported an $8.22 billion net loss for the second quarter on Wednesday, virtually all of it an unrealised markdown on the company's bitcoin holdings under fair-value accounting rules. The loss came to $24.45 per diluted share, more than eleven times the $2.19 analysts had penciled in, and marked the sharpest reversal from a $10.02 billion net income in the same quarter a year earlier.
Bitcoin holdings grew to 846,000 coins at quarter-end, an 11 per cent increase from the prior period. Strategy's blended acquisition cost sits at $75,578 per bitcoin. The reference price used for the mark-to-market at 30 June was roughly $58,700, and that gap of about $16,900 across the stack accounts for the bulk of the $8.32 billion unrealised loss the company booked against its digital assets.
The number is misleading only in the sense that it changes every quarter without any operational cause. Fair-value accounting, which Strategy adopted alongside every other bitcoin treasury company at the start of 2025, requires the entire holdings to be re-priced against market at each reporting date. When bitcoin runs, the paper gain flows through the income statement; when it drops, so does the loss. Neither corresponds to a sale.
What did happen operationally is more interesting. Strategy rebranded from MicroStrategy earlier this year and has continued to compound its treasury by issuing equity and preferred stock rather than debt. During the quarter the company raised $8.41 billion through at-the-market offerings, split $2.95 billion of common MSTR shares and $5.47 billion of STRC preferred stock. Convertible debt fell 18 per cent to $6.7 billion. The USD reserve rose 12 per cent to $2.4 billion. Bitcoin per share, the metric Michael Saylor has spent two years training the market to watch, ticked up 5 per cent.
STRC is the story inside the story. The Variable Rate Series A Perpetual Stretch Preferred Stock, launched last year as Strategy's flagship digital-credit product, has grown from a $2.8 billion notional stack at the end of Q4 2025 to $10.5 billion at the end of Q2 2026. The security now sits as the top holding in the BlackRock PFF, Virtus InfraCap PFFA and VanEck PFXF preferred stock indexes. Its effective yield sits at 13.6 per cent, or roughly 21 per cent on a tax-equivalent basis.
That yield is the transmission mechanism from the equity market into the bitcoin trade. STRC pays a floating coupon that adjusts with market rates, and the entire spread over Strategy's cost of capital is available to be recycled into more bitcoin purchases. When STRC trades below par, as it has for most of Q2, the arithmetic gets more favourable still: Strategy repurchased 288,930 shares between 20 and 26 July at an average of $86.52, roughly a 13 per cent discount to the $100 stated value. About $975 million remains available under its Digital Credit Securities Repurchase Program.
The buyback discipline is worth taking seriously. Strategy's public commitment is to be a regular and disciplined purchaser of STRC below par, with the pace increasing at deeper discounts and tapering as the security approaches $100. That is functionally the opposite of the conversion mechanics on the older convertible notes, and it removes one of the tail risks equity holders were worried about a year ago: that a prolonged drawdown would force dilutive issuance at the wrong moment.
None of this changes the underlying arithmetic of the treasury. Strategy still owns bitcoin at an average of $75,578 in a market that closed the quarter near $58,700, and every 1 per cent move in bitcoin now moves the reported income statement by roughly $500 million. The company has been trimming positions to fund preferred dividends, and any sustained period of bitcoin trading below the blended average will keep the mark-to-market losses coming.
The market read the earnings as neutral to positive. MSTR closed higher on the day of the report despite the loss, in line with the increasingly consistent pattern of the stock trading on Bitcoin per share and STRC economics rather than the reported bottom line. That is the outcome Saylor set out to engineer when the fair-value accounting change first landed; the Q2 print is the cleanest evidence yet that it has taken hold with the shareholder base.
The next test is the coming quarter. If bitcoin recovers, the Q3 income statement will flip back to a large paper gain and the loss will be forgotten. If it drifts sideways, Strategy will keep issuing STRC and buying below-par preferred, extending the transmission chain by which fixed-income capital gets converted into bitcoin at spot. If it breaks lower, the mark-to-market losses compound and the STRC yield becomes the pressure valve. All three paths are already priced in to the model.
