Capital B just bought 13 more Bitcoin. It paid €970,000. The market value of what it already held sits roughly €50.8 million below what the company paid for it. That single pair of numbers tells you more about the institutional Bitcoin treasury trade than any quarterly letter from a fund manager. The company did not buy the dip because it saw a catalyst. It bought because its financing mechanism has no off switch.
To understand why a Paris-listed entity keeps adding to a position that is underwater by a wide margin, you have to stop reading it as a bet on Bitcoin and start reading it as a structured product. The structure is an At-The-Market issuance agreement, a custody arrangement licensed under MiCA, and a shareholder register dominated by two names. Everything else is execution detail.
Context matters here because Capital B is not MicroStrategy, and the difference is not one of scale alone. MicroStrategy built its position with a layered capital stack: convertible notes priced at zero or near-zero coupons, preferred equity with defined liquidation preferences, and the occasional equity raise. That stack gives the firm optionality. It can choose when to lever, when to dilute, and how to stagger maturities against the Bitcoin cycle. Capital B runs a thinner machine. Its visible funding channel is the ATM, a rolling issuance facility that lets the company sell newly created shares into the market whenever a subscriber steps up. When TOBAM subscribed 172,978 shares at €5.68, the proceeds landed at approximately €980,000. The Bitcoin purchase of €970,000 followed within the same window. The residual gap of about 1.6 percent is close enough to be a rounding-and-fee artifact. The inference is direct: the coins were bought with the subscription money. This is not a company deploying a war chest. It is a company converting subscriptions into satoshis on a near one-to-one basis.
The custodian is Swissquote Bank Europe, operating under a MiCA license. On paper this is a meaningful upgrade over the offshore, unregulated custodian arrangements that dotted the 2021 cycle. MiCA gives the arrangement a supervisory hook, capital requirements, and a paper trail that an auditor can follow. But the chain is only as strong as its weakest node, and here the weakest node is concentration. A single custodian holds a single asset for a company whose entire equity story is that single asset. There is no multi-signature scheme, no geographically distributed sharding of keys, no self-custody fallback described in the filings. The MiCA license protects the company from regulatory action. It does not protect the company from the custodian.
Here is the part that gets buried under headline numbers. The company reports a Year-to-Date BTC Yield of 2.20 percent. That figure is not a return in euros or dollars. It is a ratio: the growth in Bitcoin held per fully diluted share. When a company issues shares to buy Bitcoin, two quantities move at once. The Bitcoin stack rises, but so does the share count. BTC Yield measures whether the numerator outran the denominator. A positive reading means the issuance was accretive to holders on a per-share basis. A reading near zero means the machine is running to stand still.
A treasury strategy funded by continuous equity issuance is a treadmill, and the treadmill only works while the asset price cooperates.
Run the arithmetic. If a company issues €980,000 of stock and buys €970,000 of Bitcoin, the per-share Bitcoin backing only improves if the new shares were sold at a premium to the existing Bitcoin-per-share value. In a rising market, that is easy. The stock trades above its net asset value, the ATM prints, the coins accumulate, and the yield metric flatters management. In a falling market, the same mechanism inverts. The stock trades at or below net asset value, the ATM dilutes holders at a worse exchange rate, and the yield metric stalls. Capital B is currently reporting a 2.20 percent YTD figure in a market where its treasury is materially underwater. That combination is the signature of a machine that is being kept warm rather than one that is compounding.
The shareholder register sharpens the picture. Blockstream Capital Partners holds 18.67 percent. Adam Back holds 17.54 percent personally. Together, roughly 36 percent of the equity sits with two related parties. That concentration is a governance fact with technical consequences. It means the ATM strategy has a reliable anchor buyer, which stabilizes the issuance channel. It also means the float available to public shareholders is thin, which can make the stock price a poor signal of the underlying Bitcoin value. When the float is thin and the treasury is a single volatile asset, the market price of the equity can diverge from its net asset value for extended periods. That divergence is the exact condition under which ATM issuance becomes dilutive to remaining holders.
Code does not lie, but it often omits the truth, and a corporate treasury is a codebase with an unaudited function — the function where the price of the new shares is decided.
The MiCA framework addressed custody, disclosure, and market integrity for crypto-asset service providers. It did not write rules for how a listed company should value a Bitcoin treasury on its balance sheet in a drawdown, nor did it define the accounting treatment when ATM proceeds are converted to coins at a loss relative to the existing basis. Those are traditional financial reporting questions, and the original report did not disclose the audit firm, the fair-value methodology, or the impairment policy. That omission is not trivial. If the treasury is carried at cost, the €50.8 million gap is an unrealized loss that sits off the income statement until impairment is triggered. If it is carried at market, the company has already booked the loss. The two treatments produce radically different equity narratives, and the public cannot tell which one is in force.
Now the contrarian angle. The prevailing read on underwater treasury companies is that they are victims of a bear market, that they simply bought too high, and that the damage is a pure price story. That read is comfortable and wrong. The damage is a structure story. A company that funded its stack with a mature, diversified capital stack — converts, preferred, staggered maturities — can absorb a drawdown because it can wait for the cycle. A company that funded its stack with a rolling ATM and a concentrated, related-party buyer base has a much shorter runway. Its ability to keep buying depends on whether subscribers keep stepping up. Its ability to survive depends on whether those subscribers have the same incentives as the public holders. When Blockstream and Adam Back together own a third of the equity, the question of who bears the dilution is not answered by the yield metric. It is answered by who gets to decide when the ATM reopens.
There is a second blind spot. The €50.8 million figure is measured in euros against a cost basis. Bitcoin does not have a cost basis. It has a spot price and a holder. A treasury that is underwater in euros is not necessarily underwater in a structural sense — if the holder can hold through the cycle, the mark-to-market loss is a financing cost, not a terminal event. The structural risk is not the drawdown. It is the combination of the drawdown with a funding mechanism that requires a healthy equity price to function. When equity trades below net asset value and Bitcoin trades below cost basis simultaneously, the ATM becomes a value-destroying channel. That is the state Capital B needs to avoid, and its 2.20 percent yield says it is not there yet.
Scalability is a trilemma, not a promise. The same logic applies to treasury companies. Capital B is trying to scale a Bitcoin position, a shareholder base, and a financing channel at once, and it cannot optimize all three. If it prioritizes the Bitcoin stack, it dilutes faster. If it protects the share count, it buys slower. If it protects the yield metric, it needs the asset price to rise. The company is currently choosing the first option, which is why the coins keep accumulating while the treasury stays underwater. That is a deliberate engineering choice, not an accident.
What to watch next is specific and measurable. First, the ATM subscription cadence. If TOBAM and related parties keep buying at current price levels, the machine stays funded. If the cadence slows, the Bitcoin accumulation stops, and the strategy loses its only moving part. Second, the disclosure of the audit treatment and the fair-value policy. Until that surfaces, the €50.8 million gap is an estimate, not a fact, and any reader treating it as a hard number is reading a headline, not a balance sheet. Third, the custodian concentration. A single MiCA-licensed custodian is better than none, but a treasury whose entire value proposition depends on one institution's operational integrity is a single point of failure with a regulatory wrapper.
The deeper question is not whether Capital B survives the drawdown. It is whether the ATM-funded treasury model survives contact with a bear market at all. In a rising tape, every one of these structures looks like genius. In a falling tape, the same structures reveal which ones were engineered to hold and which ones were engineered to keep issuing. The coins do not care. The subscribers will.