Bitcoin treasury companies: capital structure, not holdings, decides who owns the bitcoin
The seniority of the claims stacked ahead of common stock — set by both strategy and jurisdiction — matters more than how much bitcoin a company holds.
A “bitcoin treasury company” is not really an operating business; it is a machine for converting capital-markets access into bitcoin per share. In 2026 the three leading examples — Strategy (Nasdaq: MSTR), Strive (Nasdaq: ASST) and Metaplanet (TSE: 3350) — all flipped from a premium to a discount to the value of their bitcoin, and the model only compounds value above parity. Once you account for the debt and preferred stock that rank ahead of common shareholders, the apparent hierarchy inverts: the largest, most sophisticated balance sheet is not the safest, and the most-constrained company leaves common holders the cleanest claim. Figures are point-in-time as of 22 June 2026 and move weekly.
- The sector-wide premium has gone. By mid-2026 Strategy (847,363 BTC), Metaplanet (40,177 BTC) and Strive (~19,900 BTC) all trade below the value of their bitcoin, and all three sit beneath their average cost. The bull-market flywheel only spins above ~1.0× mNAV.
- Differences run on two distinct axes. Strategic: Strategy and Strive are both US-listed with the same toolkit but opposite choices — four funding rails versus one. Jurisdictional: Metaplanet operates in Japan, where the rules forbid the US preferred-stock playbook and a NISA tax shelter creates retail demand no US issuer can replicate.
- After senior claims, common owns far less bitcoin than the headline implies. “Claims drag” is ~45% at Strive, ~38% at Strategy and ~16% at Metaplanet — so preferred stock, not size, drives the drag, and the company barred from preferred (Metaplanet) gives common the cleanest claim (~84% of the bitcoin).
- Strategy's credit rail is visibly stressed. Its variable-rate preferred STRC de-pegged to ~$89 (≈11% below par) on 18 June 2026, new issuance paused, and the company made its first bitcoin sale since 2022 to fund preferred dividends.
- The decisive question for an allocator is no longer “how much bitcoin?” but “what sits in front of the common shareholder?” Scale and financial sophistication are not the same as resilience.
Investors and corporate-development leaders evaluating bitcoin-treasury equities should:
- Read mNAV and claims drag together, never holdings alone — ask what debt and preferred rank ahead of common before crediting a company with its full bitcoin reserve.
- Treat the preferred-dividend obligation (on the order of US$1.5–1.7bn/yr at Strategy) as a fixed cash liability that bites hardest precisely when markets are weakest.
- Separate strategic from jurisdictional differences: the first are changeable choices, the second are structural. Do not benchmark Metaplanet as a smaller Strategy — it is competing for a different (tax-sheltered, domestic-retail) pool of capital.
- Be willing not to pay up for scale. In a discount regime the smaller, more constrained issuer can offer the cleaner economic claim on the underlying asset.
- By end-2026, if bitcoin stays below treasury-company cost bases, sector-wide claims drag rises further as preferred stacks grow faster than bitcoin holdings, widening the gap between headline “bitcoin per share” and what common equity actually owns.
- Through 2027, “stable” variable-rate preferreds (e.g. STRC, SATA) will repeatedly trade away from par in stressed conditions, making them an unreliable funding rail below NAV and forcing issuers toward buybacks, asset sales or dormancy.
- By 2027, jurisdiction-specific structures — most visibly Japan's NISA-driven retail demand — will produce persistent valuation divergence between US and non-US treasury companies that is largely independent of bitcoin's own price.
The model: a flywheel governed by one number
Every bitcoin treasury company runs the same loop: raise capital in public markets (shares, preferred, bonds or warrants), buy bitcoin, let the larger reserve support a higher share price, and raise again more cheaply. Whether the loop helps or harms existing owners turns on mNAV — market value divided by the value of the bitcoin behind it. Above 1.0× (a premium), issuing stock to buy bitcoin is accretive: every share ends up backed by more bitcoin. Below 1.0× (a discount), the same act destroys bitcoin per share. From 2020 to 2024 the model — led by Strategy — traded at large premiums; 2026 is the test of what happens when the premium disappears.
Why now: a sector-wide regime change
Through 2025 the premium compressed, and around late 2025 Strategy crossed into a sustained discount (its first since early 2024). By mid-2026 all three names trade at a discount, and with bitcoin near US$64–65k — below every cost basis — the treasuries are also underwater on paper. Crucially this happened to all of them at once: the apparent diversification of holding several treasury companies is less real than it looks.
How it differs: two axes
Strategic (Strategy vs Strive, both US). Strategy maximises optionality — common equity, five preferred series (~US$15.5bn), convertible debt (~US$6.7bn, much at a 0% coupon) and a cash reserve — at the cost of ~US$1.5–1.7bn/yr in preferred dividends. Strive does the opposite: a single preferred instrument (SATA, a variable-rate perpetual paid daily at 13%, issued at US$90 vs a US$100 par), no debt, cleaner for common but with no fallback rail. Same market, opposite bets.
Jurisdictional (Metaplanet, Japan). Tokyo exchange rules require preferred dividends backed by recurring cash flow, so Metaplanet postponed its preferred-stock plans and funds itself through moving-strike warrants (with a built-in 1.01× mNAV accretion floor) and bonds. In return, Japan offers what the US cannot: a NISA account in which the stock is tax-free versus up to ~55% on bitcoin held directly — a structural magnet that has drawn 240,000+ retail holders.
What common actually owns (the CEBE lens)
Bitcoin per share counts the warehouse; it ignores the claims against it. Subtracting net senior claims (debt + preferred − cash, in bitcoin terms) gives the bitcoin attributable to common — and the “drag” ranks the field: Strive ~45%, Strategy ~38%, Metaplanet ~16%. Because preferred stock drives the drag, the firm Japan barred from preferred leaves common owning the most (~84%) and is the only one of the three still cheap once senior claims are stripped out. (Strategy's drag was reproduced directly from its filings; the others are filing-based estimates.)
Where it works, and where it does not
| The model rewards… | The model punishes… |
|---|---|
| A premium to NAV (mNAV > 1.0×) — issuance is accretive | A discount (mNAV < 1.0×) — issuance destroys bitcoin per share |
| Light senior claims ahead of common (low drag) | A heavy preferred/debt stack with a fixed dividend bill |
| Cheap, deep capital markets (US) or a tax-driven domestic bid (Japan/NISA) | Reliance on a single funding rail, or on retail sentiment that can reverse |
| Bitcoin appreciating above the cost of the leverage | A prolonged period with bitcoin below cost and below NAV |
Adoption and maturity
- Proven: Strategy — the scale reference case, deepest capital-markets access.
- Emerging: Strive's single-rail, debt-free design; Metaplanet's warrant-with-accretion-floor mechanism.
- Contested: Metaplanet's stated 100,000-BTC-by-2026 target (stretched in a discount); the durability of NISA-driven retail demand.
Risks and caveats
Reflexivity cuts both ways and arrived on both rails at once (equity discount + STRC de-peg). Strategy carries a dividend treadmill; Strive carries single-rail concentration; Metaplanet carries execution and retail-dependence risk. All three are underwater on bitcoin, so reported results swing with its price. CEBE/claims drag is a useful analytical methodology, not an issuer or standards-body metric — treat the ranking, not the decimal, as the signal, and re-verify against filings.
Guidance for evaluation
- Pull the latest 8-K/disclosure for current holdings, debt, preferred and cash — figures move weekly.
- Compute claims drag (net senior claims ÷ bitcoin) and read it next to mNAV.
- Identify the observed funding rail this period and whether it is open (issuing) or stalled (de-pegged/paused).
- Classify each difference as strategic (a choice) or jurisdictional (a constraint) before comparing across borders.
- Strategy Inc Form 8-K, 22 Jun 2026 — 847,363 BTC as of 21 Jun; US$64.10bn cost; US$1.4bn USD Reserve; weekly ATM/BTC update (SEC EDGAR, CIK 1050446).
- Strategy “$1.5bn Debt Repurchase / capital-structure update”, 26 May 2026 — converts ~US$6.7bn, preferred ~US$15.5bn notional; first bitcoin sale since 2022.
- STRC pricing (market reporting), May–Jun 2026 — slip below par; record low ~US$89 on 18 Jun 2026; paused issuance.
- Strive Inc Form 8-Ks, Jun 2026 — SATA 13.00% paid per business day (15 Jun); ~19,864 BTC (18 Jun); no debt.
- Metaplanet disclosures and reputable reporting, Apr–Jun 2026 — 40,177 BTC (held since 2 Apr); postponed preferred listing (Tokyo market structure); moving-strike warrants; NISA tax treatment; “555 Million Plan” targets.
- CEBE / claims-% framework — bitcoin per share after senior claims, computed from company filings; figures as of 22 Jun 2026.
ATM — At-The-Market (continuous equity/preferred issuance) · BTCTC — Bitcoin Treasury Company · CEBE — Common Equity Bitcoin Exposure · mNAV — market value to bitcoin Net Asset Value · NAV — Net Asset Value · NISA — Nippon Individual Savings Account (Japan tax-advantaged account) · SATA / STRC — variable-rate perpetual preferred shares (Strive / Strategy).