Due Diligence Capital Notes

The number itself is almost hard to process. As of May 31, 2026, Strategy Inc. holds 843,706 Bitcoin — the largest corporate Bitcoin treasury ever disclosed, accumulated at an aggregate cost of $63.87 billion. That is roughly 4% of all the Bitcoin that will ever exist, sitting on one company's balance sheet.

But here is what caught my attention in the June 1 filing: during the final week of May, Strategy simultaneously sold 801,994 shares of its own common stock for $128.3 million in net proceeds and sold 32 Bitcoin for $2.5 million. Two transactions running in parallel, each serving a completely different purpose. The share sale was accumulation fuel — capital raised to buy more Bitcoin. The Bitcoin sale was the opposite: a small liquidation to fund dividend payments on the company's preferred stock. Understanding why both activities happen in the same week, and what the coexistence of those two mechanics tells us about where this capital structure is heading, is what this post is about.

The Engine Underneath the Treasury

Strategy's core business model is simple to state and genuinely unusual in practice: raise capital through equity and structured debt instruments, deploy it into Bitcoin, repeat. But the machinery powering that loop has grown considerably more sophisticated over the past eighteen months.

As of today, Strategy operates with five concurrent preferred stock series — STRF, STRC, STRE, STRK, and STRD — alongside a massive ATM (At-The-Market) equity program that allows the company to sell common shares directly into the market in controlled daily increments, and a dedicated cash cushion the company calls the USD Reserve. Each instrument is aimed at a different investor type. Together they form a machine engineered to run in almost any market environment — pulling capital in from fixed-income allocators, yield-seeking retail investors, and index-driven equity buyers, all simultaneously.

How the ATM Program Works

ATM (At-The-Market) offering is a mechanism where a company registers a block of shares with the SEC and then drips them into the open market over days, weeks, or months — rather than doing a single large overnight offering that would immediately pressure the share price. The company controls the pace, can pause at any time, and captures the prevailing market price rather than accepting a discount.

Strategy's current ATM program is enormous by any standard. On March 23, 2026, the company staged a new $21.0 billion MSTR common stock ATM tranche. Combined with remaining capacity from the prior facility, the June 1, 2026 8-K filing discloses a total of $26.1 billion in remaining MSTR ATM capacity. At current Bitcoin prices, $26 billion buys roughly 200,000 additional BTC.

During just the May 26–31 window, the company sold 801,994 shares generating $128.3 million in net proceeds — a week's worth of quiet, incremental selling absorbed by the market without distress. This is the concept of accretive dilution in action: because Strategy's stock trades at a premium to its underlying Bitcoin value (its NAV, or Net Asset Value — the Bitcoin holdings minus debt, divided by shares outstanding), selling new shares at that premium and using the proceeds to buy Bitcoin actually increases the BTC backing per share rather than diluting it. The pie gets sliced into more pieces, but the pie grows faster than the slicing.

The 8-K is direct about the pipeline: "Strategy announced a new $21.0 billion offering of MSTR Stock (the MSTR Increase). The MSTR Stock amount available for issuance reflects the aggregate remaining capacity of both the current offering and the MSTR Increase. Sales under the MSTR Increase may begin once capacity under the existing offering is substantially depleted." That is not a company winding down its accumulation. That is a company pre-staging the next phase before the current one is finished.

The Preferred Stock Layer

The preferred stock program is where the capital structure becomes genuinely intricate. Preferred stock sits between debt and common equity: preferred holders receive fixed dividends before common shareholders see a cent, but they don't carry the same hard repayment obligation as a bond. Think of it as a structured yield product that happens to rest on top of a Bitcoin treasury.

Five series are currently active, each targeting a different buyer:

STRF, STRE, STRD: quarterly dividend of $2.50 per share
STRK: quarterly dividend of $2.00 per share, with a conversion feature — the option to exchange into common shares at a future price, giving holders some of the Bitcoin upside
STRC: a variable-rate series, currently set at 11.50% per annum effective June 1, 2026, paying a monthly distribution of $0.958333 per share

All five series declared dividends payable June 30, 2026. Running five series simultaneously lets Strategy access capital from pension allocators who want fixed quarterly income, from yield-focused retail investors who want high monthly distributions, and from more sophisticated buyers who want a hint of Bitcoin upside embedded in a preferred structure. It is a deliberately broad net cast across the fixed-income investor universe.

There is also a notable tax characteristic worth flagging: Strategy's filing states that preferred dividends are expected to be characterized as non-taxable return of capital for U.S. federal income tax purposes. Return of capital distributions effectively reduce the holder's cost basis rather than generating immediately taxable income — a meaningful advantage for taxable accounts. Whether this treatment persists long-term is not guaranteed, but it adds another layer of appeal for certain investor profiles today.

The USD Reserve: A Liquidity Buffer

The third pillar is the USD Reserve — a cash cushion that Strategy first announced on December 1, 2025. As of May 31, 2026, it stands at $900 million. The purpose, per the company's own language, is to "support the payment of dividends on Strategy's preferred stock and interest on its outstanding indebtedness" — without having to sell Bitcoin into a falling market to meet those obligations.

That design choice matters more than it might first appear. Bitcoin's worst stretches tend to coincide with broader market stress and tightening liquidity. Forced selling into that environment — selling BTC at exactly the wrong moment to cover fixed cash obligations — would be deeply destructive. The USD Reserve exists precisely to break that chain.

That said, it is not an unlimited buffer. This period's filing reveals that 32 BTC were sold at an average price of $77,135 per coin, generating $2.5 million specifically to fund preferred distributions — with the filing noting explicitly: "Proceeds from the bitcoin sales are expected to be used to fund distributions on preferred stock." So the $900 million reserve is the primary shock absorber, but small BTC sales serve as the secondary release valve when needed. As the preferred program scales, the math around this valve warrants close attention.

The Treasury in Numbers

Let me anchor the scale here because the raw figures are significant:

843,706 BTC held as of May 31, 2026 — the largest disclosed corporate Bitcoin position in recorded history
$63.87 billion aggregate purchase cost
$75,699 per BTC average acquisition price across the entire stack
$900 million USD Reserve as of May 31, 2026
$26.1 billion in remaining ATM equity issuance capacity

The average acquisition price of $75,699 is the single most important number for modeling risk here. It is the weighted anchor point of the entire treasury. At current market prices, the portfolio carries a substantial unrealized gain — but that gap can close. The capital structure was built assuming Bitcoin's long-run trajectory continues upward. If Bitcoin spends an extended period meaningfully below that average acquisition price, the entire edifice faces pressure: NAV compresses, the ATM program stops being accretive, and preferred dividend coverage increasingly depends on equity issuances at unfavorable terms.

You can review the full filing and the complete ATM and BTC update disclosures at the SEC EDGAR filing index for Strategy's recent 8-K filings.

What Could Break This Thesis

Every piece of financial engineering contains the seeds of its own undoing if the assumptions underneath it prove wrong. Here are the four failure modes I track most closely.

A sustained Bitcoin price decline below $75,699 per coin. The entire structure depends on Bitcoin's long-run appreciation. If BTC drops meaningfully below the average acquisition cost and stays there, unrealized losses accumulate, the NAV premium that makes ATM sales accretive evaporates, and the feedback loop that makes this strategy work runs in reverse. This is the primary existential risk, and it is entirely outside management's control.
Dilution from $26.1 billion in remaining ATM capacity. Accretive dilution is only accretive when the stock trades at a premium to NAV. If that premium compresses — due to a falling BTC price, a market de-rating of the structure, or loss of investor confidence — selling shares stops being value-additive and starts being punishing for existing common shareholders. Over $26 billion in potential share issuance is a large sword that swings both ways.
Growing preferred dividend obligations outpacing the USD Reserve. Five concurrent preferred series generate fixed quarterly and monthly cash obligations. Right now, $900 million plus small BTC sales manages the load. But as the preferred program grows in size — and the direction of travel is clearly toward more series, not fewer — the dividend burden compounds. A prolonged flat or declining BTC environment would force either meaningful Bitcoin liquidations or equity issuances at terms that common shareholders would find painful.
Tax treatment uncertainty on preferred dividends. The non-taxable return-of-capital characterization of preferred distributions is currently expected, not guaranteed. A change in IRS guidance, or a company-level income event that shifts the characterization, could increase effective tax costs for preferred holders and reduce demand for future preferred issuances — making that channel of capital more expensive.

Conclusion

What Strategy has assembled is not really a technology company in any traditional sense. It is a financial architecture — one specifically designed to route capital from the global equity and fixed-income markets into a single, supply-capped, non-yielding asset, at scale and with regularity.

The June 1 filing is a weekly status report on that architecture: 843,706 BTC secured, $128.3 million in equity raised in a single week, five preferred series serviced, a $900 million liquidity buffer running, and $26 billion in future equity capacity pre-staged and waiting. Every moving part is doing exactly what it was designed to do.

The question I keep returning to is not whether the design is clever — it clearly is. The question is whether Bitcoin's long-run appreciation rate stays comfortably above the blended cost of all this capital: the preferred dividends at 11.50% and above, the ongoing dilution embedded in ATM sales, the interest on outstanding debt. If it does, common shareholders end up owning a concentrated, compounding position in the world's hardest asset through one of the most sophisticated accumulation vehicles ever built. If it doesn't, the leverage works against them with the same efficiency it works for them on the way up. That asymmetry is the entire thesis, laid bare in a weekly 8-K.