Last week a headline crossed my feed: Strategy Inc. had overtaken First Abu Dhabi Bank by market capitalization. I read it three times, looking for the numbers. There were none. No mNAV. No BTC per share. No price, no timestamp, no primary source — no SEC filing, no exchange notice, no terminal print. Six claims, four of them unsourced assertions, two of them opinion, and a quantitative spine made entirely of air.
I have learned to treat that absence as the finding. Based on my audit experience — two hundred hours spent mapping voting centralization in Compound's governance mechanism in the summer of 2020 — I know that the missing column of a balance sheet usually explains more than the columns that were printed. A company measured by its market capitalization while its BTC per share goes unmentioned has not been reported on. It has been priced by mood.
The structure underneath deserves a serious reading, and it gets too little of it. Since 2020, Strategy — formerly MicroStrategy — has operated as a Bitcoin treasury company: a listed equity whose core asset strategy is accumulating BTC, funded less by operating cash flow than by capital markets. The mechanism is a reflexive loop. When the shares trade above the per-coin net asset value of the holdings — when mNAV exceeds one — the company issues equity at a premium, converts that premium into coins, lifts BTC per share, and reinforces the narrative that sustains the premium.
Soros called this reflexivity. I call it a structure with no interior. Fundamental value here is not produced; it is asserted, then arbitraged. Faith in people is costly; faith in math is free. And a market capitalization ranking contains almost no math. In a market that chops sideways, that distinction matters more than usual: consolidation is where positioning gets built, and positioning built on a misread denominator is positioning built on sand.
Market capitalization is the wrong denominator. For a treasury vehicle, the number that matters is BTC per share — total holdings divided by diluted shares — and its companion, mNAV, the multiple the market pays over the per-coin net asset value. Those two figures tell you whether the flywheel is compounding or quietly eroding. Market cap tells you what a sentiment coefficient looked like at a moment that has already passed. When the premium compresses toward one, issuing shares to buy coins dilutes holders rather than enriching them, and the loop inverts. The inversion is not a risk scenario; it is the arithmetic of the structure.
The comparison is also a category error. A bank under Basel capital rules is priced against booked equity, regulated capital ratios, deposit franchise and credit pricing power. A treasury vehicle is priced against BTC, leverage, and the market's willingness to pay a premium over liquid, near-zero-fee substitutes. Placing them on one axis produces a headline, not an insight. I reviewed more than forty whitepapers during the 2017 cycle for a series I called The Hollow Promise, and the recurring trick was identical: choose the benchmark that flatters the claim, then let the benchmark do the arguing.
The competitor is not a bank. It is the ETF. A spot Bitcoin ETF offers the same exposure linearly, at a fraction of the fee, with no premium to expire and no convertibles to refinance. Every basis point of fee compression inside that wrapper is a basis point of justification removed from the premium vehicle. That is the compression story — the slow, structural one that never makes a headline, because nothing dramatic happens on any single day.
The financing toolbox is a ladder, and both ends are sharp. ATM equity issuance dilutes at whatever premium exists when the printer runs. Convertible notes defer dilution but create a cliff: above the conversion price, conversion delivers a concentrated share supply; below it, the obligation is repaid in cash or refinanced. Preferred shares sit senior to common and carry rigid dividend duties. Layered together, they buy time — and time is precisely what a leverage structure purchases when the underlying asset is volatile.
Accounting has stopped being neutral. Under FASB ASU 2023-08, crypto assets are marked to fair value through net income starting with fiscal 2025. The old impairment model let paper losses hide. The new one makes quarterly earnings swing with the BTC chart — and earnings are what debt covenants and rating agencies read. A company can be operationally sound and still breach a covenant because a weekend candle moved. We audit the logic, for humans will always err; the ledger, at least, does not flatter.
Then there is the governance question nobody asks. A listed vehicle built around a decentralized asset is governed by a conventional board, with founders holding voting power well beyond their economic stake. Capital allocation — when to issue, when to buy, at what pace — concentrates in a handful of hands. Shareholders hold no vote on the execution that defines the thesis. I spent a Berlin roundtable with twelve women building in the NFT space, listening to how little leverage they had over the platforms that carried their work, and the pattern rhymed: the asset is decentralized; the decision is not.
Custody is the unmentioned seam. Large institutional holdings typically sit with a small number of qualified custodians, supplemented by self-custody and multisig arrangements. That is a single-point counterparty exposure sitting underneath an asset whose entire premise is removing counterparties. Code is the only law that does not sleep — but code governs the chain, not the vault, not the insurer, not the prime broker. The custody architecture is disclosed in fragments; the media never asks, because the media is busy comparing market caps.
And the cohort is growing. Treasury-company imitators now exist across Tokyo, New York and Hong Kong. Individually, each is a choice. Collectively, they are a correlated bid that becomes a correlated offer when funding windows narrow. The same reflexivity that amplifies accumulation amplifies liquidation. Index inclusion adds a second mechanical layer: passive inflows on the way up, forced deletions on the way down. A structure that borrows the market's enthusiasm also inherits its mechanics.
Here is where I argue against my own instinct. The prevailing read is that this milestone proves crypto has arrived. The contrarian read is narrower and, I think, more useful: a treasury vehicle is not a technical achievement at all. It is a financial engineering structure, and its steadiest beneficiaries are the underwriters — the banks arranging the notes, the desks running the ATM programs, the lawyers drafting the indentures. They collect fees on the way up and on the way down. The people selling shovels are not the people holding the pickaxe.
There is an honest steelman, though, and I will give it room. In jurisdictions where a spot ETF is unavailable, a listed treasury company is a genuine access primitive, and for an investor who understands they are buying volatility amplification rather than a Bitcoin substitute, that can be an intentional position. My objection is not to the instrument. It is to the story, which converts a leveraged proxy into proof of victory and asks no one to read the covenant, the maturity table, or the disclosure.
What I will watch is not the ranking. It is the mNAV multiple, the BTC per share trend across two consecutive quarters, the convertible maturity ladder, fee compression inside the ETF wrapper, and index reconstitution dates. Those are measurable; a headline is not. Where the premium holds, the loop compounds. Where it decays, the same loop distributes losses to the last buyers of the narrative. Hype burns out; robustness remains in the ledger. The ledger has not yet been shown to us — and until it is, I will keep reading the filing, not the feed.