
With premiums compressed, the metrics investors use to evaluate bitcoin-heavy balance sheets have not kept up. Strategy's filings show the gap: realized losses funding dividends, preferred equity distorting NAV calculations, and no standard for pledged coins.
For most of the last two years the digital asset treasury conversation has focused on the wrong things: which company announced the largest bitcoin purchase, which executive made the boldest promise about never selling, and how high a premium the market would assign to a balance sheet full of bitcoin. That framing worked during an accumulation cycle. It no longer works now that the premiums have compressed, and in many cases inverted.
Two consecutive filings from Strategy illustrate the point. During the week of August 3 through August 9, the company sold 1,690 bitcoin for roughly $109 million and used the proceeds to repurchase its STRC preferred stock. The following week it sold no bitcoin at all. It raised $334 million by issuing common stock, allocating $52 million to preferred dividends and $132 million to further share repurchases. The remaining $149 million went into its dollar reserve. The obligations stayed the same while the funding source shifted in under a week. No headline metric currently in circulation captures that shift, according to analysts who follow the sector.
Multiple on net asset value, or mNAV, is not a GAAP measure and was never designed to function as one. It is generally calculated as enterprise value divided by the dollar value of a company's digital asset holdings, with enterprise value including market capitalization plus debt and preferred equity. That final component is where the issues begin, because preparers and data providers do not apply it consistently. Preferred equity has become a meaningful part of these capital structures, not a rounding error.
Bitcoin per share carries the same problem. Twenty One Capital's second quarter report, filed with the SEC on August 11, disclosed 12,547 satoshi per Class A share as of June 30, down from 12,557 at the end of 2025. The stated objective is to grow holdings faster than the share count. Over the first half of 2026 the company did not. The filing attaches its own warning label, noting that the figure measures neither the ability to service liabilities nor book value per share.
The company holds 43,514 bitcoin worth roughly $2.8 billion while the equity trades near $1.6 billion, which works out to about 57 cents paid for every dollar of bitcoin inside. Adjust that calculation for debt and cash, as Bitcoin Treasuries does, and the same position prints at 0.70 times. Same balance sheet, same day, with two defensible answers roughly thirteen percentage points apart. Neither carries a footnote explaining which convention was used.
While the investor-facing metrics smooth volatility, the financial statements do not. Strategy reported a net loss of $8.2 billion for the second quarter of 2026, driven by an $8.3 billion unrealized loss on digital assets, against quarterly revenue of $122 million from its software business. That is the fair value model working exactly as the standard setters intended. It produces income statements in which the operating company is nearly invisible.
The realized side also deserves attention. Strategy has sold 6,948 bitcoin for approximately $432 million since May, beginning with 32 coins in its first disposal since 2022. Recent sales priced near $64,262 against an average purchase price of $75,385. Those are realized losses funding a perpetual dividend, a materially different fact pattern from an unrealized markdown that may reverse.
Management at both companies has been unusually candid about the dynamic. Strategy's Phong Le has observed that preferred holders prioritize cash liquidity over bitcoin holdings. Twenty One's Rapha Zagury opened his first shareholder letter by agreeing with the company's critics. He wrote that the gap between market value and bitcoin held could be viewed as a misallocation of capital and that management shares that view. Those are honest concessions about what a capital structure demands once the premium stops doing the work. The reporting framework has not caught up to them.
The first obstacle is terminology and classification. Until there is agreement on which share count and which components of enterprise value belong in these calculations, comparability across companies is illusory. The sector now includes well over a hundred public issuers of very different sizes.
The second issue is how these cryptoassets are treated when pledged as collateral. Bitcoin per share treats every coin as equivalent, but 16,116 of Twenty One's coins, more than a third of the stack, sit pledged against $485 million in convertible notes. The company holds a cash position of $106 million. A metric that cannot distinguish between a free coin and a pledged one is not a liquidity measure.
The third item is governance. The figure that actually matters is the least glamorous one: months of dividend and interest coverage held in cash. Strategy's dollar reserve reached $4.8 billion as of August 16, which management characterizes as roughly 2.7 years of obligations. Boards and audit committees should be asking when routine asset sales to service a perpetual obligation stop being a capital allocation decision. That question will only become more important from here.
Prepared with AlphaScala editorial tooling from the source reporting linked above. Indexable analysis may include a cited Alpha Score value. Publishing checks screen each story before release. Educational coverage, not personalized advice.