I spent a long weekend with a spreadsheet I did not enjoy building. At its center sat a number that has been circulating through my Nairobi study group for weeks: 847,666. That is the reported size of Strategy's Bitcoin treasury โ the largest single corporate position in the world โ and a figure my younger students cite the way my generation once cited the Federal Reserve's balance sheet. What unsettled me was not the size. It was the second number I typed beside it: a minus sign in front of the company's own "BTC Yield" metric. Strategy is now buying more Bitcoin than ever while its shareholders quietly own less of it per share. That sentence sounds like a contradiction. It is arithmetic, and it deserves to be read slowly.
I have spent most of my adult life auditing the space between what a system claims and what it actually does. In 2017, reviewing ERC-20 proposals in this same city, I learned that a token standard can be technically neutral and socially distorted at the same time โ the bias hides in edge cases nobody markets. The same instinct returns here. For weeks the financial press has framed Strategy's resumed accumulation as a triumph, a signal that the largest institutional believer in Bitcoin is doubling down. The headline is not false. It is also incomplete, and the missing part is exactly where the risk lives.
Strategy โ the company formerly called MicroStrategy โ has built something genuinely novel, and I want to honor that before I interrogate it. Over four years it has assembled a machine that turns capital-market instruments into Bitcoin. The engine has three moving parts. The first is an at-the-market (ATM) common-stock program, which still holds roughly $18.84 billion of unused issuance capacity; last week it sold about 1.47 million shares for $246.2 million. The second is a matrix of preferred stock: STRC, STRF, STRK, and STRD, each carrying a fixed or variable dividend, with STRC paying a striking 12% annualized. The third is a cash reserve โ roughly $5 billion set aside specifically as a USD buffer, plus about $1 billion in general cash โ designated for servicing those dividends. On paper, the machine converts equity and preferred capital into an appreciating hard asset. The company's total holdings now sit near 847,666 BTC, purchased at an average cost of roughly $75,437 per coin, implying an unrealized gain of about 10.6% at current marks.
That is the official story, and on its own terms it is coherent. What breaks my calm is the interaction between two metrics the company publishes side by side but rarely reconciles out loud.
The first metric is absolute holdings. By this measure, Strategy is a spectacular success: a company that keeps buying, quarter after quarter, converting financial paper into digital scarcity. Its most recent disclosure reported a net quarterly increase of roughly 1,666 BTC โ a modest but symbolically important return to accumulation after a period of net outflow.
The second metric is BTC Yield, the company's own invention. It measures not how much Bitcoin Strategy owns, but how much Bitcoin each share represents. And here the number turns cold: year-to-date BTC Yield is approximately negative 3.7%, a BTC Gain of minus 25,085 coins, translating to a shortfall of roughly $2.1 billion. For the quarter, the negative BTC Gain reaches roughly minus 100,275 coins, or about $8.37 billion in value terms.
These two numbers do not contradict each other. They simply describe different things, and the gap between them is the entire story. When a company issues new shares or preferred stock faster than it accumulates Bitcoin, the absolute pile grows while every existing shareholder's slice of that pile shrinks. This is dilution wearing the costume of growth. Strategy's treasury expanded; its shareholders' proportional claim contracted. The company has spent four years teaching the world to watch the Bitcoin, and in doing so it has trained us to ignore the denominator.
This is where my audit reflex sharpens. A balance sheet is a moral document as much as a financial one โ it tells you who is being enriched and who is being asked to wait. The standard defense is that Strategy only issues when its market value exceeds the value of its Bitcoin holdings, meaning each new share is accretive. That defense depends entirely on something called mNAV: the market-to-net-asset-value premium. While that premium is high, issuing shares genuinely adds per-share value. When it compresses toward zero, the same issuance math reverses. Here is the quiet alarm: back-solving last week's numbers โ roughly $246.2 million raised across about 1.47 million shares โ implies an MSTR share price near $167. If that is accurate, the premium that justifies the whole flywheel may already be thinning, and the accretion argument weakens with it.
I want to be careful, because the reported figures carry a caveat I must flag honestly. One disclosure in the source material carries a 2026 date that does not align cleanly with the public record, and the 847,666 BTC figure is larger than what I can independently confirm. I would not build a position on any of this without cross-checking Strategy's SEC 8-K filings directly. The integrity of an argument depends on the integrity of its inputs, and no analysis โ mine included โ is better than the data beneath it.
Still, the internal logic of the disclosed strategy is legible, and it points in one direction. Pay attention to where the money actually went last week. Analysts had expected Strategy to deploy somewhere between $250 million and $300 million on Bitcoin. It deployed $142.7 million โ roughly half. Meanwhile, it spent $151.7 million repurchasing its own STRC preferred stock, slightly more than it spent on Bitcoin itself. When a company redirects more capital to defending the price of its financing instrument than to acquiring the asset it exists to accumulate, something has shifted. The buying engine is no longer pointed primarily at Bitcoin. It is pointed at keeping the funding channel intact.
That pivot makes sense the moment you look at STRC's mechanics. The instrument trades near a $100 par value, and it recently climbed from a low around $71 back to roughly $99. That looks like a recovery. But of the volume in STRC, Strategy itself accounts for an estimated 18.9% โ down from 22.8% the prior week, but still enormous. A market price that is largely set by the issuer buying its own instrument is not a market price; it is a managed number. The move back toward par is better read as the visible result of a $2 billion buyback program (already more than $1 billion spent) than as independent demand returning. If the buyback pauses, the support beneath that price pauses with it.
There is a deeper, colder mechanism here that I have not seen named clearly enough in the commentary, and it is what worries me most. The plan to shift STRC โ along with STRF, STRK, and STRD โ from monthly payments to daily distributions, pending a shareholder vote, is presented as an enhancement. Engineers will recognize it for what it is: a duration play. Paying daily shortens the effective duration of the preferred instrument, reducing discount pressure and nudging STRC closer to a money-like product. The real purpose is to lower the cost of defending the price. It is a defensive optimization disguised as an upgrade, and I say that with respect for its cleverness rather than contempt for it.
Step back and the architecture of the whole model becomes visible. Strategy does not generate operating income. Its returns flow from two sources only: Bitcoin appreciating, and its securities selling at a premium that lets it raise capital accretively. That is not a Ponzi scheme โ there is real Bitcoin backing it, a genuine asset on the ledger, which distinguishes it from the frauds I have spent years helping communities avoid. But it is a reflexive structure in the precise sense that the philosopher of markets George Soros meant: its own success is a precondition for its continuation. Rising Bitcoin supports the premium; the premium funds more Bitcoin; the cycle reinforces itself. Reflexive cycles are beautiful on the way up and merciless on the way down, because the same loop runs in reverse.
Here is the contrarian point, and I want it heard as an observation rather than a prophecy. The universal reading of Strategy's resumed buying is that institutional conviction has returned. The subtler reading is that the buying volume halved against expectations while capital migrated toward propping up the funding instrument, and that the flagship per-share metric turned negative. When the loudest story in crypto is accumulation, the quiet story is often dilution โ and the market's attention always runs ahead of its arithmetic. The company's decision to hold $6 billion in reserve against the $8.37 billion scale of its negative quarterly BTC gain is not the behavior of an institution that believes the loop is frictionless. It is the behavior of one that has already priced in the possibility that it may need to pay dividends for a long time without buying much Bitcoin at all.
Where does this leave the honest observer? I think of the small libraries my colleagues and I built across Kenya during the DeFi Summer, translating mechanics into Swahili so that people could see the structure of the thing for themselves rather than trust the price. That work taught me that education is not a hedge against risk โ it is the only honest relationship we have with it. Reading Strategy well does not require optimism or pessimism. It requires keeping three numbers in view at once: the average cost of its Bitcoin, near $75,437; the per-share BTC Yield, now negative; and the mNAV premium that decides whether issuance helps or harms. Trace those three together and you will know the truth of the model long before the press release does. The silence between the blocks is where the signal hides. Somewhere a shareholder is watching the price rise and the slice shrink, and no rally will ever tell them the difference. Watch the denominator. It always speaks last.