Strategy bought 267 bitcoin. It also spent $193 million repurchasing its own STRC perpetual preferred stock and lifted the repurchase authorization from $2 billion to $3 billion. Only one of those numbers tells you where this cycle actually sits. It is not the one that made the headline.
At $100,000 a coin, 267 BTC is roughly $27 million — a rounding error against a treasury of 848,517 BTC. The buyback is seven times larger. And the price action that likely forced it is the most under-reported line in the whole release: STRC was rejected at $100 par and closed below $98.
A par-anchored perpetual preferred — monthly distributions, no maturity date, distributed largely through offshore retail channels — failed to hold the single number that defines its product identity. That isn't a footnote to a bitcoin story. That's the credit market settling its position in cash while equity analysts debate whether an index headline is priced in.
I'll flag the data before I build on it. The 848,517 BTC figure sits roughly 200,000 coins above the disclosure sequence I tracked through late 2025, and the MSCI consultation on "non-operating companies" publicly maps to October 2025, not the later timestamp circulating with this release. I take the numbers as stated. I hold the absolute values at arm's length. Twenty-three years of watching this industry has taught me that the discrepancy is usually in the timestamp, and the timestamp is usually the tell.
What STRC actually is
Strip the branding and the instrument is a traditional perpetual preferred: $100 par, monthly dividend, no redemption date, sold to yield-seeking buyers. The pitch calls it digital credit. The mechanics call it a money-market substitute with a leveraged bitcoin balance sheet behind it. Retail distributors in Asia and the Middle East do not buy it because they want bitcoin beta. They buy it because it pays a coupon and looks like cash.
That is the entire design premise. Par is not a valuation for this instrument. Par is its identity. Once STRC prints $97.80, the buyer who thought he owned a dollar-denominated yield product discovers he owns a duration instrument with an equity cushion nobody modeled. Distribution channels do not renegotiate. They stop calling.
Meanwhile the rest of the capital structure is a genuine feat of financial engineering. Common equity absorbs the beta. Convertibles go to arbitrage funds that only care about realized volatility. STRF carries a fixed coupon for rate-sensitive income buyers. STRC floats. The ATM sells common into premium. Each tranche targets a different buyer with a different risk appetite and a different exit. It is, in the most literal sense, composability — packaging one volatile asset into four instruments with four distinct risk profiles. Composability isn't a philosophical trap here; it's a maturity mismatch wearing a governance costume.
The arithmetic nobody led with
Run the two disclosures side by side. JPMorgan's estimate of $2.8 billion in potential passive outflow if MSCI excludes the company against roughly $27 million of marginal buying. That's a ratio near 100 to 1. The bid is a rounding error next to the hypothetical offer.
Now run the liquidity math, which the release doesn't do at all. $5.3 billion in cash and equivalents against a bitcoin position worth roughly $85 billion at current marks. Coverage under 7%. The software business is too small to service the preferred dividends and convertible interest. Refinancing is the funding source, not operations. Every quarter that the financing window narrows, the coupon gets paid out of the balance sheet rather than out of cash flow.
And the buyback itself: when a par product trades below par, repurchasing it is the only mechanism that defends the unit of account. You cannot defend a peg with marketing. You defend it with cash. The company spent seven times more defending its liability than expanding its asset. That sequencing — preferred first, bitcoin second — is more informative than any public statement about long-term conviction.
To be precise about the structural accusation that always follows this company: it is not a Ponzi. There is no promised fixed return to equity holders, the assets are real and audited, and there is no closed loop of new money paying old interest. What exists is reflexivity. Share premium feeds financing capacity, financing capacity feeds coin accumulation, coin accumulation feeds the narrative that sustains the premium. That loop runs in both directions at the same speed. I made this exact argument during the Terra forensics in May 2022 — the tell was never the headline number, it was the shape of the redemption curve three days before anyone else charted it. The curve here is STRC at $98.
The blind spot in "priced in"
"Already priced in" is the most dangerous sentence in this entire episode. It presupposes the market has correctly forecast the final scope of the index decision. The list lands in three weeks. Uncertainty is not a price. It is an unresolved variable, and unresolved variables get resolved by whoever writes the definition.
That definition is the real weapon. MSCI does not need to prove a violation. It only needs to describe the business differently — from operating company to investment vehicle. Saylor's response was that it was disappointing. No legal challenge, no procedural objection. That silence is itself a data point: the classification is not contestable on the merits, because it is not a finding of fact. It is a taxonomy choice. A closed-end fund loses the tech multiple, loses the index slot, loses the passive bid. One reclassification, three losses.
Here is the part that should keep holders up at night. The company's success is eroding the identity that justifies its valuation. The larger the treasury, the more the entity resembles a fund, and the more defensible the non-operating label becomes. Scale is not neutral here. Scale is the evidence.
And the competitive threat was never the other treasury companies. It is the spot ETF sitting one click away, offering bitcoin exposure with no leverage, no governance concentration, no refinancing risk, at a few basis points. Every month the premium compresses, that substitution gets cheaper.
Don't wait for the final list to form a view. The preferred market has already voted, in cash, below par, before the committee met.
What to watch next
The number that matters this quarter is not the coin count. It is the STRC spread to par, published daily, free, and ignored. Second: the composition of the next raise. Common sold into a premium means the loop still turns. Preferred sold into a discount means the loop is being fed from the expensive end.
Investors who wait for the index decision will be reading a verdict on a question the credit market answered weeks ago. The open question is narrower and harder: when the premium compresses and the retail preferred channel closes in the same quarter, what exactly is left holding the 4% of global supply?