Due Diligence Capital Notes

Every week, like clockwork, Strategy sells a portion of its own shares to the public, takes the cash, and immediately buys Bitcoin. Last week was no different: 297,940 shares of MSTR common stock sold into the open market, $39.7 million collected, 592 Bitcoin added to the corporate treasury. The total stack now sits at 717,722 BTC, acquired for approximately $54.56 billion in aggregate since the company began its accumulation strategy.

What makes this worth examining isn't the round numbers — it's the machinery underneath. Strategy isn't funding Bitcoin purchases with operating profits from its software business. It isn't drawing on a revolving credit facility or negotiating a new convertible bond every week. It's running a continuous, low-friction equity-to-Bitcoin conversion engine using a financial structure called an ATM program — and it maintains five of them simultaneously across five different Nasdaq-listed securities. That's a capital-recycling flywheel I haven't seen replicated at this scale anywhere else in public markets.

What Is an ATM Program, and Why Does It Matter Here?

An ATM offering — "at-the-market" in full — is a mechanism that allows a public company to sell newly issued shares directly into the open market, a small batch at a time, at whatever the current trading price happens to be. Unlike a traditional secondary offering, which involves hiring underwriters, announcing a fixed share count, and pricing a deal overnight, an ATM is more like a tap. The company opens the tap slightly, shares flow out into the market, cash flows in, and existing shareholders are diluted by a small fraction. The tap closes. The process repeats on whatever schedule the company chooses.

ATM offerings aren't new — real estate investment trusts and other capital-hungry companies have used them for decades to fund property acquisitions. What Strategy has done is industrialize the concept across an entire ecosystem of securities. Instead of running one ATM program for its common stock, the company maintains five concurrent programs across five different instruments, each targeting a different category of investor. The combined remaining capacity across all five programs stands at approximately $37.35 billion — meaning Strategy could, in theory, raise that much more capital without returning to its investors for a new mandate. That number alone deserves a moment of pause.

The consequence of this structure is that Bitcoin accumulation stops being a discrete corporate event and becomes a continuous background process. There's no press conference, no analyst call, no announcement. Just a regulatory 8-K filing, usually on Monday or Tuesday, telling you what happened during the prior week.

The Five-Security ATM Stack: How Each Piece Works

Strategy maintains five Nasdaq-listed securities, each with its own ATM program and its own natural investor audience. Here's how the stack is constructed:

MSTR Common Stock (Class A) — This is the core vehicle: traditional equity in Strategy Inc. Shares sold via this ATM program fund Bitcoin purchases directly. As of February 22, 2026, the remaining capacity in this program stands at $7,843.9 million. The 297,940 shares sold during the February 17–22 week generated $39.7 million in net proceeds — the exact capital used to purchase that week's 592 Bitcoin. No other program was tapped during this period.
STRK (8% Perpetual Preferred Stock) — This security pays an 8% annual dividend and includes a conversion feature allowing holders to exchange it for MSTR common shares at a predetermined future price. It appeals to income-oriented investors who want some Bitcoin upside but need a predictable yield stream as a floor. The remaining ATM capacity here is enormous: $20,331.6 million, making STRK the single largest untapped reserve in the entire structure.
STRF (10% Perpetual Preferred Stock) — A fixed 10% annual dividend with no conversion feature. Closer in investor psychology to a corporate bond — you're buying the yield, not the equity option. Remaining capacity: $1,619.3 million.
STRC (Variable Rate Perpetual Preferred Stock) — Unlike STRF and STRK, STRC pays a floating dividend rate, offering some protection for investors concerned about rising interest rates. Remaining capacity: $3,542.8 million.
STRD (10% Perpetual Preferred Stock) — The newest addition to the preferred stack, also carrying a 10% fixed rate. Remaining capacity: $4,014.8 million.

Taken together, the four preferred programs retain approximately $29.5 billion in unused capacity. Not one of those four programs sold a single share during the February 17–22 week. Strategy drew only from the common stock ATM — a small, deliberate pull from a very deep well.

Reading the February 23 Filing

The 8-K filed with the SEC on February 23, 2026 gives a precise view of how this machine operates in practice. The filing contains two disclosure items: Item 7.01, a Regulation FD notice pointing investors to Strategy's public dashboard at strategy.com, and Item 8.01, the weekly Bitcoin update.

The Bitcoin update is where the operative data lives. During the week of February 17–22, Strategy acquired 592 BTC at an all-in average cost of $67,286 per coin, inclusive of fees and expenses, for a total outlay of approximately $39.8 million. The filing is explicit about the funding source: "The bitcoin purchases were made using proceeds from the sale of shares under the ATM."

Here is the detail that I find most worth flagging. Strategy's cumulative all-in average cost basis across all 717,722 Bitcoin is $76,020 per coin — meaning the company has paid an average of $76,020 for every coin it has ever purchased, from its very first buy to last week. But this week's 592 coins came in at $67,286 — roughly $8,700 below that blended average. Every Bitcoin bought during this period pulled the company's cost basis downward, not upward. In portfolio accounting terms, the company bought at a discount to its own book value per coin. The filing notes that "aggregate and average purchase prices are inclusive of fees and expenses" — so there's no hidden cost being excluded from that comparison.

The filing was signed by Thomas C. Chow, Strategy's Executive Vice President and General Counsel. That's a standard formality for 8-K submissions, but it's worth recognizing: this is a routine, board-authorized capital allocation activity, filed on the same schedule every week.

The Flywheel Logic

Why does this mechanism keep working? The short answer is that Strategy's stock consistently trades at a NAV premium — meaning the stock price is higher than the per-share value of the Bitcoin the company holds, net of liabilities. NAV, or net asset value, is simply what you'd get if you divided the total Bitcoin value minus total debt by the number of shares outstanding.

Because the stock commands this premium, every share Strategy sells through the ATM raises more dollars than the underlying Bitcoin backing per share would imply. Those excess dollars flow directly into purchasing Bitcoin, increasing the per-share Bitcoin backing, which in turn supports ongoing investor interest in a premium valuation. It's a reflexive loop — and like all reflexive loops, it can run in either direction.

The existence of $29.5 billion in preferred stock ATM capacity that went completely untouched last week suggests Strategy is managing the sequencing with some deliberateness. The preferred programs appear to be held in reserve for larger-scale capital raises or for periods when equity market conditions make preferred issuance particularly attractive. The common stock ATM handles the weekly rhythm.

What Could Break This Thesis

No flywheel runs forever without friction. Here are the specific scenarios that would invalidate Strategy's accumulation model:

Bitcoin staying below $76,020 for an extended period. The company's all-in average cost basis is $76,020 per coin. A sustained Bitcoin price below that level puts the entire treasury underwater on a mark-to-market basis. That doesn't automatically trigger a liquidity crisis, but it eliminates the accretive logic of selling equity at a NAV premium to buy Bitcoin — and it would test the patience of preferred shareholders whose dividends are implicitly backed by appreciating collateral.
Preferred dividend obligations outrunning available cash flow. STRF carries a 10% annual dividend, STRK carries 8%, STRD carries 10%, and STRC pays a variable rate. These aren't discretionary — preferred dividends must be paid regardless of Bitcoin's performance or how much the software business earns. If capital markets tighten and Strategy can no longer sustain ATM issuance at scale, it needs another reliable cash source to service those obligations. That source isn't clearly visible in today's balance sheet.
Dilution outpacing per-share BTC accumulation. The ATM model creates value for shareholders only when the NAV premium is high enough that each share sold buys more Bitcoin than the share previously represented in BTC terms. If that premium collapses — as it did during Bitcoin's extended 2022 bear market — the dilution math stops being accretive and starts being destructive. Shareholders end up holding a smaller percentage of a stagnant or shrinking asset base.
Forced or distressed liquidation. With 717,722 BTC on the balance sheet, Strategy holds approximately 3.4% of Bitcoin's entire circulating supply. Any forced selling event — whether triggered by margin calls, regulatory action, or refinancing pressure — would be large enough to affect Bitcoin's market price at the precise moment the company most needs the price to hold. The concentration risk runs in both directions.

Conclusion

What I find structurally significant about this week's filing isn't the size of the purchase — 592 coins is a rounding error relative to a 717,722-coin stack. It's the operational regularity of it. Strategy has built a capital allocation machine that converts equity market appetite into Bitcoin accumulation on a weekly cadence, with five distinct ATM programs offering five different entry points for five different categories of investor. The $37.35 billion in combined remaining capacity represents a standing, pre-authorized mandate to continue doing exactly this.

Whether that mandate proves wise depends entirely on Bitcoin's long-term trajectory relative to a $76,020 blended cost basis. If Bitcoin compounds above that level over the coming years, this will have been one of the most consequential corporate treasury decisions in financial history. If it doesn't, the preferred dividend obligations and dilution arithmetic will tell a very different story in retrospect. That's not ambiguity — that's the actual bet. Understanding that the bet is this explicit and this mechanical is, I think, the most useful thing any serious analysis of Strategy can offer.