"What if Strategy can't keep raising money?" is the question. The answer is in their cash flow statement.
27 months
of breathing room
As of December 31, 2025, Strategy held $2.30B in cash and short-term investments against $1.02B of annual non-BTC obligations (preferred dividends, interest, taxes, debt service, operating burn). That's 2.3 years of all recurring obligations covered without raising another dollar. Strategy's 10‑K formally names this pile the "USD Reserve" — established December 2025 — and explains it covers preferred dividends and interest on outstanding debt.
"What if you can't keep issuing shares?"
Strategy's playbook depends on continuous capital raising. The ATMs sell common stock; the preferreds (STRK, STRF, STRD, STRC, STRE) issue perpetual yield to bondholders; the converts borrow at low rates against the BTC stack. Every week a new tranche, every quarter another billion in dividends and interest to fund.
The skeptical question writes itself: what happens if the markets close? If MSTR's premium to NAV evaporates? If preferred buyers stop showing up? If a credit downturn locks the converts? Can Strategy still pay its obligations without new issuance?
It's a fair question to ask once. The USD Reserve is the answer — a deliberately accumulated cash buffer specifically for this scenario. The size and source of the buffer are in the 10‑K's Consolidated Statement of Cash Flows; the buffer policy is in MD&A; and the math is small enough to fit on one screen.
From operating-month-to-month to a $2.3B reserve
Each row is from the corresponding 10‑K's Consolidated Statement of Cash Flows. Annual non-BTC obligations grew with the preferred stack (each new preferred class adds a coupon stream); the cash buffer grew faster. The Reconcile page has the full breakdown by source and use.
| Period | Cash on hand | Annual non-BTC obligations | Buffer |
|---|---|---|---|
| FY2023 | $48.7M | $71.5M | 0.7 years (8 months) |
| FY2024 | $39.9M | $258.7M | 0.2 years (2 months) |
| FY2025 | $2,303.3M | $1,023.7M | 2.3 years (27 months) |
FY2023 had a 0.7-year buffer because the obligation stack was tiny — no preferreds yet, only convertible interest. FY2024 dropped to 0.2 years because the cash got redeployed aggressively into BTC during the post-election rally. FY@fiscalYearLabel jumped to 2.3 years because the preferred dividend stream became material AND Strategy started deliberately retaining cash from the raises rather than spending all of it on BTC. That's the buffer policy in action.
Three layers of defense, in order
The USD Reserve is the first layer. Here are the next two.
- 1. USD Reserve covers 2.3 years. Even if the equity markets locked tomorrow and Strategy couldn't issue another share, every preferred dividend and every interest payment is covered through 2027 at the current obligation rate. No new issuance, no asset sales, no behavior change required.
- 2. Preferred dividends are non-cumulative and discretionary. STRF, STRC, STRD, STRC, and STRE are all perpetual preferreds — Strategy CAN skip a dividend payment without triggering a default. The shares lose voting privileges and the preferreds become more expensive to issue going forward, but skipping is legally available. Most operators would never use this lever because it's a credibility cost, but it's there if needed.
- 3. BTC is liquid and unencumbered. Strategy holds 700K+ BTC entirely outside the indebted entities; the BTC is not pledged as collateral. In a doomsday scenario where layers 1 and 2 are exhausted, selling 0.1% of holdings covers a year of obligations at current BTC prices. This is the explicit answer to "but BTC could go to zero" — even a 99% BTC drawdown still leaves enough holdings to cover several years of obligations.
None of these are how Strategy plans to operate; they're insurance layers. Layer 1 (the buffer) covers the realistic case where capital raising temporarily slows. Layer 2 covers a more severe disruption. Layer 3 is the tail-risk hedge nobody expects to need.
"Permanent" raise capacity is not a hope
Strategy is a publicly listed company on Nasdaq. Equity issuance via ATM is permanent infrastructure for a public company — the question is never whether they can issue, only at what price the market clears. Even in a severe drawdown, ATM activity continues at smaller volumes; the share count goes up by less, but it goes up.
Conditional pause scenarios — "if MSTR trades below NAV" or "if the premium disappears entirely" — would slow down the rate of new preferred issuance and make convertible borrowing more expensive. They don't make capital raising impossible. The S&P 500 weathered five drawdowns of 20%+ since 2020; public companies kept raising capital through every one of them.
The honest framing of "what if they can't raise?" is not "what if it becomes impossible" — that's not a real outcome. The honest framing is "what if it becomes more expensive?" That's a real risk and one that affects the growth rate of the BTC stack, not the survival of the obligation stack. The USD Reserve is sized for the second framing; the first framing isn't a real scenario.
Defensive optionality, not necessity
The USD Reserve is not a hedge against catastrophe. It's a pre-positioned response to a question the markets keep asking — "what's your buffer?" — that Strategy didn't have a clean answer to until December 2025. Now they do. 2.3 years of all non-BTC obligations, sitting in cash and short-term investments, available immediately.
Whether you think the reserve is too small, too large, or just right is a separate conversation about appetite. What it is not is a sign of fragility. A company without a buffer can be fragile; a company that built a multi-year buffer specifically for the failure scenario is the opposite of fragile. Saylor has spent 12 years anticipating critiques. This is the anticipation in the cash flow statement.
Full annual breakdown of how the cash flowed in and out is on the Reconcile page. Audit trail showing every weekly tranche reconciles to the SEC filings is on the Ledger page. Why we trust the BTC count itself is on the Custody page.