Data note: The latest tranche shown below is from (14 days ago). Strategy's purchase cadence is roughly weekly; if newer disclosures exist this site may be lagging the upstream refresh. Pages remain accurate for the period covered.
A perpetual obligation, a finite cost

An obligation that never ends. A BTC stack that doesn't either — if BTC outruns the coupon. This is that "if" made concrete.

The finite BTC cost
of a perpetual coupon

Strategy's preferreds promise their dividend forever — no maturity, no return of principal, just a fixed dollar coupon paid into the indefinite future. That sounds like an unbounded liability. It isn't. As long as Bitcoin's CAGR beats the coupon rate (adjusted for how often it's paid), the total BTC ever sold to fund those dividends converges to a finite number — and the rest of the BTC stack survives indefinitely. This page shows the math, two ways: the simple once-a-year approximation, and the cadence-correct version Strategy actually faces in practice.

First, the unit

CAGR — Compound Annual Growth Rate

CAGR is the annualized rate at which a quantity has compounded over time. If something started at value A and ended at B after N years, its CAGR is (B / A)1/N − 1. Plain language: "what constant yearly growth rate would produce the same start-to-end multiplier?"

Bitcoin started at fractions of a cent in 2009 and has compounded over the ensuing 17 years. Its CAGR depends on which window you measure:

  • Since 2013 (~$13 → ~$108K): ≈ 70%/yr CAGR
  • Since 2017 (~$1,000 → ~$108K): ≈ 70%/yr
  • Since 2020 (~$8,000 → ~$108K): ≈ 60%/yr
  • Trailing 5 years (~$10K → ~$108K): ≈ 60%/yr
  • Site default: 30% — half of recent realized, well below long-run history

The 30% default is deliberately conservative. We're projecting forward, not backward; using a number well below realized history leaves room for BTC's growth rate to slow as the asset matures, and still keeps the math working. Slide the BTC CAGR input below to see how the answer changes as you become more or less optimistic. There's nothing precious about 30% — what matters is whether your assumed CAGR exceeds the threshold for the specific coupon you're modeling.

The framing nobody uses

Income with an obligation

When Strategy raises a billion dollars selling a preferred, that's income. Real cash, on the books, free to spend on anything — buy BTC, pay opex, retire debt, whatever. The site models it as "buying BTC then selling some of it back to fund the coupon," but that's just a useful narrative. Mechanically, it's cash in.

The piece nobody talks about: that income comes with an obligation — a perpetual dividend stream owed to whoever holds those preferred shares. So the right question isn't "did Strategy raise a billion dollars?" It's "what's the value of the obligation that came with it?"

Saylor says Strategy doesn't sell BTC to pay dividends — and on a per-event basis, that's true. The model on this site treats every dividend as the equivalent of selling BTC anyway, because dilution-at-mNAV produces the same per-share outcome and the math is cleaner that way. Why the model diverges from Saylor's framing →

In dollars: infinite
11.5% on $100 par is $11.50/share/year. It never stops — these preferreds have no maturity. Multiply by ∞ years and the obligation in nominal dollars is infinite. Even by year nine of an 11.5% coupon you've cumulatively paid more than the $100 par back to holders. By year fifty it's over $575 in nominal-dollar payments per share. Looking at it this way, the company has already given back everything it raised — and is still paying. The accounting for "did the raise create value?" looks bleak.
In BTC: finite
But the company doesn't owe BTC — it owes dollars, and BTC is what they hold. As BTC compounds, the BTC needed to fund a fixed-dollar coupon shrinks every year. Sum a shrinking infinite series, get a finite number. At a 30% BTC CAGR and an 11.5% coupon, the entire forever-stream of obligations costs about 43% of the original BTC stack (under monthly cadence — the simple annual approximation says ~50%, which we'll explain below). The rest stays on the balance sheet permanently. The same coupon stream that looks crushing in dollars converges in BTC.

That divergence — infinite USD obligation sitting on top of a finite BTC obligation — is the entire reason the model works. Everything else on this page, and on the per-tranche pages, is just measuring how big that BTC number actually is.

Start concrete (the simple version)

"I hold 1 BTC. I owe 11.5% per year. BTC's CAGR is 30%."

Year 0 the coupon costs 0.115 BTC — that's 11.5% of V, paid at year-0 BTC prices. Year 1, BTC is worth 30% more (because we're assuming a 30% CAGR), so the same dollar coupon costs only 0.115 / 1.30 ≈ 0.0885 BTC. Year 2, less again. Every year the BTC bite shrinks because the denominator keeps growing.

Add up the infinite series of shrinking payments and it converges to D · V · (1+CAGR) / CAGR = 0.115 · 1 · 1.30 / 0.30 ≈ 0.498 BTC. That's the total BTC ever sold — about half of V. The other half funds the coupon forever with room to spare.

This is the simple version. It assumes the entire year's coupon is paid in one lump at the start of the year, before any BTC growth has happened. Real preferreds don't work that way — STRC pays monthly, STRK/STRF/STRD/STRE pay quarterly. The cadence-correct math gives a slightly different (and lower) number. We'll show both side-by-side below.

The two rates

Coupon bleeds. CAGR heals. Compare them.

Two rates, both per year, both as a percentage. That's the whole comparison.

Coupon (D) — the bleed
Each year you owe the coupon rate D as a percentage of face value. For STRC that's 11.5%. It never goes away. These preferreds have no maturity date.
CAGR — the heal
Each year BTC compounds at its CAGR. We use 30% as a forward-looking long-run guess. The same dollar coupon costs less BTC every year as long as CAGR is positive.

Year over year, the BTC needed to cover the coupon is divided by (1 + CAGR). At CAGR = 30%, that's ÷ 1.30 — every year's BTC bite is about 23% smaller than the year before. That's the geometric shrink that makes the infinite sum finite.

The ratio you actually care about is D vs CAGR:

  • CAGR > D — heal beats bleed. BTC sold per year shrinks fast enough that the cumulative total converges. You pay the coupon forever from a slice of your stack.
  • CAGR ≈ D — break-even. The cumulative line creeps up forever but very slowly. Tight.
  • CAGR < D — bleed wins. The series still shrinks each year (as long as CAGR > 0), but not fast enough. Cumulative total eventually exceeds your principal — you go broke.

The whole point: the obligation is infinite in time, but the BTC ever drained is finite in total. That's the trick. People conflate "perpetual coupon" with "infinite cost." They're not the same thing when the unit of account (BTC) keeps appreciating at a positive CAGR.

The picture

Bars shrink. Cumulative line levels off.

Each bar is one year's BTC sold to cover the coupon. The green line is cumulative BTC sold. The amber dashed line is your BTC principal (V). If the green line stays under the amber line forever, the coupon is funded in perpetuity.

Try a preferred:
Cadence:
0.8250.5500.2750.000V = 1.000Σ = 0.43405101520253035404550BTCYEARS
Sustainable forever
The cumulative BTC sold asymptotes to 0.434 BTC. That's less than your 1.00 BTC principal — 0.566 BTC lives forever.
Year by year

Watch the bar shrink

Year BTC sold this year Σ BTC sold BTC remaining
0 0.0000 0.0000 1.0000
1 0.1000 0.1000 0.9000
2 0.0770 0.1770 0.8230
3 0.0592 0.2362 0.7638
4 0.0455 0.2818 0.7182
5 0.0350 0.3168 0.6832
6 0.0269 0.3437 0.6563
7 0.0207 0.3645 0.6355
8 0.0159 0.3804 0.6196
9 0.0123 0.3927 0.6073
10 0.0094 0.4021 0.5979
Two formulas, side by side

Why the simple version misses

Whenever BTC's CAGR beats the coupon rate, the series converges to less than V. Heal beats bleed. The simple formula and the real-cadence formula agree on the headline conclusion (sustainable forever); they disagree slightly on the exact size of the BTC ever sold. Both numbers matter, and the gap is part of the education.

Simple (annual annuity-due)
Σ = D · V · (1 + CAGR) / CAGR
Assumes the whole year's coupon is paid in one lump at the start of each year, before any BTC growth has happened. Overstates burn because real coupons aren't paid that way. Break-even: CAGR > D / (1 − D).
Cadence-correct (ordinary annuity)
Σ = D · V / (k · ((1+CAGR)1/k − 1))
Each payment lands at the end of its period, after that period's BTC growth has happened. k is payments per year (4 quarterly, 12 monthly, 24 semi-monthly). This is what Strategy actually faces. Break-even: CAGR > (1 + D/k)k − 1.

What the simple version misses: it pretends the whole year's coupon shows up at year-0 prices. But Strategy doesn't pay anyone in January and then go silent until next January — they pay quarterly or monthly, and each payment captures BTC growth that's already happened during the period. Earlier payments get less BTC growth before being made; later payments get more. Summing up the actual schedule gives a different (smaller) total than collapsing it all to the start.

Concretely, at CAGR = 30%, D = 11.5%, V = 1 BTC, the lifetime BTC sold under each convention:

  • Annual annuity-due (simple closed form): 0.498 BTC (50% of V) — overstates by ~13% vs monthly
  • Annual ordinary (k=1, payment at year end): 0.383 BTC (38%)
  • Quarterly (k=4, STRK/STRF/STRD/STRE): 0.424 BTC (42%)
  • Monthly (k=12, STRC today): 0.434 BTC (43%) — page default
  • Semi-monthly (k=24, hypothetical STRC future): 0.436 BTC (44%) — barely changes
  • Continuous limit (k → ∞): 0.438 BTC (44%)

Notice the cadence numbers all cluster between 0.42 and 0.44 — the gap between quarterly and continuous is only ~3%, far smaller than the gap between any of them and the simple annual-annuity-due figure (0.498). The annual annuity-due simplification is the one that's noticeably off; cadence-detail past quarterly barely matters.

Sustainability thresholds for each preferred at its actual cadence: STRK at 8% quarterly needs CAGR > 8.24%; STRF/STRD/STRE at 10% quarterly need CAGR > 10.38%; STRC at 11.5% monthly needs CAGR > 12.12%. All three are well below our 30% default (and well below BTC's realized history). Try the preset and cadence buttons above to see how each preferred's math shifts.