Publicly traded companies now hold over 1.26 million BTC — more than 6% of Bitcoin's fixed 21-million supply — yet the corporate treasury model that produced this accumulation is fracturing. Strategy (MSTR), the company that pioneered the digital asset treasury (DAT) approach in 2020, trades at 0...
"Very soon there are going to be more AI agents than humans making transactions. They can't open a bank account, but they can own a crypto wallet." — Michael Saylor, Executive Chairman, Strategy (formerly MicroStrategy), Q2 2026 Earnings Call
Publicly traded companies now hold over 1.26 million BTC — more than 6% of Bitcoin's fixed 21-million supply — yet the corporate treasury model that produced this accumulation is fracturing. Strategy (MSTR), the company that pioneered the digital asset treasury (DAT) approach in 2020, trades at 0.60x its net asset value as of mid-August 2026, down from a peak of 3.4x in November 2024. Combined market capitalization of listed Bitcoin treasury firms has contracted from $134 billion to approximately $72 billion since October 2025, a $62 billion wipeout.
At the same time, the model continues to attract new entrants. Roughly 200 public companies have adopted some form of Bitcoin acquisition strategy as of July 2026. On August 18, Japan's Metaplanet announced a $134.6 million deal to take a 95.7% stake in Nasdaq-listed Super League Enterprise, creating a US-listed Bitcoin treasury vehicle called Superplanet. The bifurcation is clear: the strategy's pioneers are bleeding value while imitators continue to pile in.
Corporate buyers absorbed 166,984 BTC in H1 2026 — more than double the 81,153 BTC that miners produced in the same period. The demand-supply imbalance persists, but the financing structures underpinning it are under visible stress.
The numbers are large. As of August 2026, the top corporate Bitcoin holders by BTC count:
| Company | BTC Held | USD Value (approx.) | Avg. Cost Basis | |---------|----------|---------------------|-----------------| | Strategy (MSTR) | 843,775 | $63.5B | ~$75,419 | | Twenty One Capital (XXI) | 43,514 | $3.16B | N/A | | Metaplanet (3350.T) | 43,000 | ~$3.2B | ~$97,000 | | MARA Holdings | 35,577 | ~$2.7B | Varies | | GameStop (GME) | 4,710 | ~$368M | ~$106,000 |
Strategy alone accounts for 4.03% of Bitcoin's total supply. The top five corporate holders control roughly 970,000 BTC, or 4.6% of supply.
In Q2 2026, Bitcoin treasury firms purchased approximately 110,000 BTC — the largest quarter of corporate buying on record, according to KuCoin data. For H1 2026 as a whole, corporate buyers absorbed BTC at 2.06 times the rate of new mining output.
The 174 publicly traded companies tracked by Bitwise held 1.18 million BTC as of May 12, 2026, representing 5.35% of Bitcoin's fixed supply, valued at approximately $80.55 billion at the time.
The premise of the DAT model was that investors would pay a premium for leveraged Bitcoin exposure through public equity. That premium has inverted.
Strategy's mNAV (market capitalization divided by net asset value of Bitcoin holdings) stood at 0.60–0.68x in August 2026. The stock traded at $93–$95 per share, implying a market capitalization of roughly $33 billion against Bitcoin holdings worth over $63 billion. Investors can now buy Strategy shares and receive Bitcoin exposure at a 32–40% discount to spot.
This discount reverses the reflexive flywheel that powered the model's growth. When mNAV exceeds 1.0x, the company can issue equity at a premium, use the proceeds to buy Bitcoin, and increase BTC per share — making the stock more attractive and enabling further premium issuance. Below 1.0x, the mechanism works in reverse: equity issuance is dilutive, debt refinancing becomes expensive, and the company shifts from accretive accumulator to leveraged holder with a $1.5 billion annual carry cost.
Twenty One Capital, backed initially by Tether and SoftBank, debuted on the NYSE in 2026 and also traded below its NAV. The company holds $106.1 million in cash against $484.5 million in convertible notes and earns no revenue. Tether bought out SoftBank's stake in May; founder Jack Mallers resigned as CEO in July.
Bloomberg reported on July 22 that Bitcoin's roughly 50% decline from its October 2025 record of approximately $126,000 had "crushed companies that stockpiled tokens."
The DAT model's stress test produced a list of companies that sold Bitcoin under pressure:
Strategy: Sold 3,588 BTC in July 2026 to fund preferred share dividends. The sale broke Strategy's longstanding "never sell" narrative and, according to Seeking Alpha analysis, marked a "regime change" in the company's operating model.
MARA Holdings: Sold 23,093 BTC for $1.63 billion during H1 2026, receiving an average price of $70,631 per coin. Holdings fell from 53,822 BTC ($4.7B) at year-end 2025 to 35,577 BTC ($2.1B) by June 30. The company broadened its treasury policy to allow sales from existing holdings, not just newly mined coins. In early August, MARA pledged 18,750 BTC as collateral for $600 million in loans from Coinbase Credit and Two Prime Lending, partly to fund a $1.5 billion acquisition of Long Ridge Energy in Ohio.
Satsuma Technology: Shareholders voted to sell all 668 BTC (~$43.5M) and delist from the London Stock Exchange.
Sequans Communications: Sold 1,025 BTC and disposed of nearly 80% of remaining holdings to repay convertible debt.
Genius Group: Sold its final 84 BTC in early 2026 to repay debt.
Bitdeer: Fully liquidated 943 BTC in February 2026 to fund a pivot toward AI data center infrastructure.
CoinDesk reported on July 24 that the pattern extended further, with multiple companies selling Bitcoin to "repay debt, fund operations, finance buybacks or strengthen cash reserves." Forbes in March described the structural risks as companies faced the mismatch between Bitcoin's non-yield-bearing nature and the cash obligations of convertible notes and preferred shares.
Against this backdrop of stress, Metaplanet is doubling down. On August 18, the Tokyo-listed company announced it would contribute 2,100 BTC (~$132.1 million) plus $2.5 million in cash to Super League Enterprise (SLE) in exchange for 44,859,400 shares of common stock at $3.00 per share, plus preferred stock and warrants.
Key terms:
Metaplanet targets 100,000 BTC by end of 2026 and 210,000 BTC by end of 2027. The 2,100 BTC contributed to Superplanet represents approximately 4.9% of its current 43,000 BTC holdings.
The strategic logic is geographic diversification: a dual listing on the Tokyo Stock Exchange and Nasdaq gives Metaplanet access to both Japanese and American capital markets for future fundraising. Whether this justifies the structure depends on whether Superplanet can maintain an mNAV above 1.0x — a condition that Strategy, once the model's champion, has failed to sustain.
BTC Yield measures the percentage change in Bitcoin holdings per diluted share over a given period. It was designed to demonstrate that convertible debt issuance and equity dilution were accretive to shareholders on a per-share Bitcoin basis.
Metaplanet's BTC Yield trajectory tells a story of diminishing returns:
| Period | BTC Yield | |--------|-----------| | Q1 2025 | 95.6% | | Q2 2025 | 129.4% | | Q3 2025 | 33.0% | | Q4 2025 | 11.9% | | Q1 2026 | 2.8% |
The metric collapses as the share base expands and Bitcoin prices decline. At 2.8% quarterly BTC Yield, Metaplanet is adding marginally more Bitcoin per share, but the rate of improvement has fallen by 97% from its Q2 2025 peak.
Strategy faces a similar dynamic. With mNAV below 1.0x, any new equity issuance would be dilutive on a BTC-per-share basis. The BTC Yield engine requires a premium to operate. Without it, the metric becomes an accounting artifact rather than a growth indicator.
Value The Markets reported that both Strategy and Metaplanet had shifted their public messaging to emphasize BTC Yield over absolute accumulation targets — an implicit acknowledgment that the raw purchasing pace is no longer sustainable under current market conditions.
The capital structure risk is concrete. Strategy's CFO has flagged that if MSTR trades below $183 by September 2027, when the next convertible note put date arrives, the company would need to deploy "alternative levers" — including Bitcoin sales, refinancing, or additional STRC (Strategy's preferred share class) issuance.
The math is straightforward:
Bitcoin does not generate cash flow. Every dollar of debt service must come from equity issuance, debt refinancing, or asset sales. When mNAV is below 1.0x, equity issuance is dilutive. When Bitcoin prices are declining, refinancing terms deteriorate. The remaining option — selling Bitcoin — reduces the very asset base that defines the company's value proposition.
Twenty One Capital's position is structurally weaker: $484.5 million in convertible notes against $106.1 million in cash and no revenue. With Mallers gone and Tether now the controlling shareholder, the company's strategic direction remains uncertain. New leadership has pledged a "fuller strategy update before year-end."
Despite the stress, new entrants continue to adopt the model:
GameStop held 4,710 BTC as of May 2026 and implemented a covered-call strategy, writing short-dated options on its holdings with Coinbase as counterparty. The company pledged 4,709 of its 4,710 coins to Coinbase, which can rehypothecate the collateral — changing the accounting treatment from direct holdings to a receivable.
Q2 2026 saw 110,000 BTC in corporate purchases, the largest quarter on record, indicating that the aggregate buying trend persists even as individual companies exit.
Approximately 200 public companies had adopted some form of Bitcoin treasury strategy by July 2026, up from 174 tracked by Bitwise in May.
The contradiction is structural: the model's aggregate demand continues to absorb more Bitcoin than miners produce, while individual firms at the model's center face financing stress. New entrants benefit from lower average cost bases (having bought at lower prices than earlier adopters like Strategy), but face the same fundamental challenge — Bitcoin generates no yield, and leverage requires servicing.
The DAT model is, at its core, an arbitrage on market structure. It converts equity and debt capital into Bitcoin holdings, and relies on public market premiums to make the conversion accretive. The economic value flows to:
When mNAV falls below 1.0x, the value transfer reverses. Shareholders bear the dilution from any capital raise, while debt holders retain their claim on the company's Bitcoin. The model becomes extractive rather than accretive for equity investors.
The $62 billion decline in combined market capitalization represents value destruction concentrated among equity holders who bought at premium mNAV levels. Convertible note holders, by contrast, retain their conversion rights and downside protection.
The Bitcoin treasury model is not dead, but it is no longer self-reinforcing. The premium-to-NAV arbitrage that made the strategy accretive for early adopters has evaporated for the largest practitioners. Strategy, the sector's standard-bearer, now trades at a persistent discount to the value of its own Bitcoin — a condition that transforms every capital market action from growth-positive to dilutive.
The model's continued adoption by new entrants, and the record Q2 2026 corporate buying pace, suggest that the thesis — Bitcoin as a superior reserve asset — retains institutional appeal. But the financing structures layered on top of that thesis have proven fragile. Companies that borrowed to buy Bitcoin at $100,000+ now face debt service obligations while their collateral has declined roughly 40%.
What remains is a bifurcated landscape. Companies with low leverage, low cost bases, and operational cash flow (or access to patient capital markets) can sustain Bitcoin holdings indefinitely. Companies that relied on the reflexive premium to service debt are being forced to sell the very asset they were designed to hold. The model works until the premium disappears. For the sector's largest players, it has.