Due Diligence Capital Notes

Seven trading days. $1.28 billion raised. 17,994 Bitcoin purchased. That's not a quarterly result — that's what Strategy Inc. accomplished in a single week, between March 2 and March 8, 2026.

I've been watching Michael Saylor's accumulation cadence for a while now, and even by his own standards, the velocity has become striking. Most companies spending over a billion dollars in a week would be executing a merger or a major acquisition. Strategy is buying a commodity — the same one it bought last week and the week before — with capital it raised by selling its own stock and preferred shares through an automated market facility. What makes this worth dissecting isn't just the scale. It's the machinery behind it, and the quiet upgrade that machinery received in the same 8-K filing that announced the purchase.

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

Before getting into the specifics, it's worth nailing down a term that drives almost everything Strategy does: ATM offering, short for At-The-Market offering. Unlike a traditional secondary offering — where a company announces a large share sale at a fixed price, often at a discount, and the stock tanks on the news — an ATM lets a company drip-sell shares into the open market continuously, at prevailing prices, through a designated broker-agent. No discount, no single announcement to spook the market, no wall of supply hitting at once.

For a company like Strategy, which trades at a significant NAV premium (meaning its stock price is worth considerably more than the per-share value of its Bitcoin holdings), an ATM is a financial gift that keeps giving. Every share sold at a premium above the underlying Bitcoin value raises more capital than the Bitcoin is worth at that moment. The company then deploys that capital to buy more Bitcoin, which mechanically increases the Bitcoin backing per share. This is what Strategy calls BTC Yield — and it's the reason issuing new shares doesn't necessarily hurt existing holders the way standard dilution logic would suggest. Done at a premium, it can actually increase what each remaining share represents in Bitcoin terms.

Strategy currently runs this machine across five separate securities at once: Class A common stock (MSTR), and four series of perpetual preferred stock — STRK, STRD, STRC, and STRF. Each security has its own ATM program, its own broker-agent arrangement, and its own pool of remaining issuance capacity.

Breaking Down the March 2–8 Purchase

Here's exactly what happened in that one-week window, pulled directly from Strategy's Form 8-K filed March 9, 2026:

Bitcoin purchased: 17,994 BTC at an average price of $70,946 per coin.
Total cost: approximately $1.28 billion.
Funding source: entirely from ATM proceeds. As the filing states plainly: "The bitcoin purchases were made using proceeds from the sale of shares under the ATM."

Breaking that $1.28 billion down by security:

MSTR Class A common shares: 6,327,541 shares sold, generating $899.5 million in net proceeds.
STRC Variable Rate Preferred shares: 3,776,205 shares sold, generating $377.1 million in net proceeds.

So roughly 70% came from common stock sales and 30% from the STRC preferred series. The total net proceeds were $1,276.6 million — close enough to $1.28 billion that the rounding barely matters.

After this purchase, Strategy's aggregate Bitcoin holdings stand at 738,731 BTC with an all-in average acquisition price of $75,862 per BTC, representing a cumulative spend of $56.04 billion. That's not a typo. The company has deployed fifty-six billion dollars into a single asset since 2020.

The Quiet Infrastructure Upgrade: Dual-Agent Sales Outside Market Hours

Buried near the back of the same 8-K is a detail I find more interesting than the purchase itself. Strategy amended its Omnibus Sales Agreement — the master legal contract that governs its multi-security ATM program — to allow something it couldn't do before.

Previously, Strategy could only use one agent per security class on any given trading day during regular market hours (9:30 a.m. to 4:00 p.m. ET). The amendment, effective March 9, 2026, adds the following carve-out, quoted directly from the filing:

"the requirement under the Sales Agreement that Strategy will only sell securities of any single class or series through one agent on any single trading day shall not apply to or prohibit Strategy from appointing a second agent to effect sales of such class or series under the Sales Agreement before 9:30 a.m. and/or after 4:00 p.m. New York City time on such trading day."

In plain terms: Strategy can now run two broker-agents simultaneously on the same security — one during regular hours, one during pre-market or after-hours sessions. This expands the daily execution window from 6.5 hours to effectively the full pre/during/post-market window.

Why does this matter? Two reasons. First, it increases the raw volume Strategy can move on any given day. If the BTC opportunity is there and the preferred stock is bid well in after-hours, Strategy no longer has to wait for the next morning's open. Second, it's an operational signal: the company is actively engineering its capital-raise infrastructure to be faster and more continuous. This isn't a one-time upgrade — it's a sign that the playbook is being optimized for higher throughput.

The Remaining Runway: $29.6 Billion Still to Deploy

One question that comes up frequently is whether Strategy is running out of ammunition. The short answer: not yet.

Here's the remaining ATM issuance capacity across each security as of the March 9 filing:

STRK (8% fixed annual dividend): $20.33 billion remaining
STRD (10% fixed annual dividend): $4.01 billion remaining
STRC (variable rate): $3.16 billion remaining
STRF (10% fixed annual dividend): $1.62 billion remaining

Combined, that's over $29.6 billion in authorized but unissued capacity — and that number doesn't include future shelf registration renewals or the common stock ATM, which still has substantial room.

The preferred stock architecture is worth pausing on. STRK, STRD, STRC, and STRF each carry a fixed or variable dividend yield — 8% to 10% annually — and are perpetual, meaning they have no maturity date and the company has no obligation to ever redeem them at par. For investors in these preferred series, the attraction is a relatively high, predictable yield backed (indirectly) by a Bitcoin-heavy balance sheet. For Strategy's common shareholders, raising capital through preferred stock is potentially better than issuing more common shares, because preferred dividends are fixed obligations rather than a claim on upside.

Strategy also operates a real-time public dashboard at strategy.com as a Regulation FD-compliant disclosure channel — meaning it treats that website as an official source of material information alongside SEC filings. The filing explicitly describes it as covering "market prices of its outstanding securities, bitcoin purchases and holdings, certain key performance indicator metrics and other supplemental information."

What Could Break This Thesis

I want to be direct about the scenarios that would cause real damage here, because the structure creates genuine vulnerabilities alongside the opportunities.

A sustained Bitcoin price decline below $75,862. That's Strategy's all-in average acquisition cost per BTC. Under fair-value accounting, sustained prices below that level would generate significant unrealized losses and potential impairment charges. The company doesn't have an operational cash flow engine to absorb those losses — its economics are almost entirely a function of BTC price.
Dilution accumulating faster than BTC appreciation. ATM issuances work for existing shareholders only as long as the NAV premium remains high and BTC keeps rising. If the premium collapses — say, because a competitor offers cheaper Bitcoin exposure — issuing new shares at or below NAV would be destructive rather than accretive. With over $29.6 billion in remaining ATM capacity, the dilution potential is not theoretical.
Preferred dividend obligations becoming a cash drain. Four series of perpetual preferred stock, yielding 8%–10% per year, create fixed annual cash obligations entirely disconnected from Bitcoin's price. If BTC falls sharply and Strategy's common stock collapses with it, the preferred dividends still come due. The company would need to raise cash by selling Bitcoin at depressed prices or issuing more shares into a weak market — an ugly feedback loop.
Execution risk in off-hours trading. The new dual-agent amendment allows selling before 9:30 a.m. and after 4:00 p.m. ET. Pre-market and after-hours sessions typically have wider bid-ask spreads and thinner liquidity. Selling large blocks of preferred stock or common shares in those windows could result in worse execution prices than during regular hours, quietly eroding the proceeds available for each Bitcoin purchase.

Conclusion

What Strategy did in the week of March 2–8, 2026 is a data point that should recalibrate how investors think about the company's operational pace. This isn't a firm that dips into the capital markets occasionally to buy Bitcoin opportunistically. It has built a multi-security, multi-agent, always-on capital raising infrastructure — and it just upgraded that infrastructure to operate outside regular market hours.

The 738,731 BTC on the balance sheet, accumulated at an all-in cost of $56.04 billion, is the score so far. But the more important number might be the $29.6 billion in remaining ATM capacity that hasn't been deployed yet. If Bitcoin's trajectory over the next few years even partially justifies Saylor's conviction, that remaining runway represents an enormous lever. If it doesn't — if BTC stalls or declines while preferred dividends compound and common shares dilute — the same machinery works brutally in reverse.

I'm watching the pace of accumulation, the NAV premium, and the Bitcoin price relative to that $75,862 cost basis. Those three numbers, tracked together, tell most of the story.