The Digital Asset Treasury (DAT) model — the corporate finance innovation that promised to turn publicly traded companies into leveraged crypto proxies — is breaking apart in real time. Strategy Inc. (formerly MicroStrategy) now trades at a 30% discount to the net asset value of its 714,644 Bitco...
"If bitcoin falls 90% for the next four years, we'll refinance the debt. We'll just roll it forward." — Michael Saylor, Executive Chairman, Strategy Inc., CNBC Squawk Box, February 10, 2026
The Digital Asset Treasury (DAT) model — the corporate finance innovation that promised to turn publicly traded companies into leveraged crypto proxies — is breaking apart in real time. Strategy Inc. (formerly MicroStrategy) now trades at a 30% discount to the net asset value of its 714,644 Bitcoin, ETHZilla has collapsed 98% from its August 2025 peak, and Peter Thiel's Founders Fund has dumped its entire stake. Meanwhile, MARA Holdings quietly moved $87 million in Bitcoin to counterparties in a single day as its stock cratered 35% in a month.
What began as a financial innovation — using corporate balance sheets to offer regulated Bitcoin exposure before spot ETFs existed — has devolved into a structural trap. More than 200 U.S. companies have announced plans to adopt digital asset treasury strategies, seeking to raise approximately $102 billion to buy crypto for their corporate treasuries. Most will never get there. The DAT model requires perpetually rising crypto prices, cheap capital markets access, and investors willing to pay a premium for the middleman. In February 2026, all three conditions have evaporated.
Strategy Inc. holds 714,644 Bitcoin acquired for an aggregate cost of approximately $54.35 billion — an average price of $76,056 per coin. At Bitcoin's current price near $66,000, those holdings are worth roughly $47 billion. The company's market capitalization sits at approximately $43 billion.
This is the core indictment of the DAT model: Strategy's stock now trades at a 0.7x price-to-NAV ratio. Investors can buy $1 of Bitcoin exposure through MSTR for 70 cents — and they're still choosing not to.
The premium-to-NAV was the entire economic engine. When MSTR traded at 2x or 3x its Bitcoin holdings' value, Saylor could issue convertible bonds at 0% interest, sell shares via at-the-market offerings, and use the proceeds to buy more Bitcoin — creating a reflexive loop where buying drove the premium, which enabled more buying. That flywheel has reversed. The stock has fallen 66% over six months, erasing nearly $90 billion in market capitalization.
The structural reason is straightforward: spot Bitcoin ETFs, launched in January 2024, eliminated Strategy's reason to exist. Investors no longer need a middleman to access Bitcoin exposure. BlackRock's iShares Bitcoin Trust (IBIT) offers direct, low-cost Bitcoin exposure without the corporate overhead, debt obligations, or management risk. Strategy's premium compressed from its peak above 3x to below 1x — a permanent regime shift, not a temporary dislocation.
Yet Saylor continues buying. Between February 9 and February 16, Strategy acquired 2,486 Bitcoin for $168.4 million at an average price of $67,710. The conviction is undeniable. Whether it's brilliance or recklessness depends entirely on where Bitcoin trades in 2027, when the first $1.2 billion convertible note matures.
If Strategy is the stress-tested original, ETHZilla is the cautionary tale of what happens when the model gets copied without the conviction.
ETHZilla launched in August 2025, raising $565 million to build an Ethereum treasury on the MicroStrategy playbook. Peter Thiel's Founders Fund took a 7.5% stake, triggering a 90% single-day surge that pushed the stock to an effective peak of $174.60. Sixty institutional investors piled in, including Electric Capital, Polychain Capital, and GSR.
Fourteen months later, ETHZilla trades at $3.62 per share — a 98% collapse. On February 18, 2026, an SEC filing confirmed that Thiel's Founders Fund now holds exactly zero shares, having exited the entire position during Q4 2025.
The failure was structural, not just market-driven. When Ethereum's price declined sharply, ETHZilla did the opposite of what the MicroStrategy model demands: it sold. In October, the company liquidated approximately $40 million in ETH for stock buybacks. In December, it sold another 24,291 ETH worth $74.5 million to repay debt. Its holdings fell from over 100,000 ETH at the peak to 69,802 ETH today — making it only the sixth-largest corporate Ethereum holder.
The capitulation confirmed the market's worst fear about DAT copycats: they lack the conviction to hold through drawdowns. Strategy has never sold a single Bitcoin in five years. ETHZilla sold within months. That behavioral asymmetry destroyed credibility and, with it, any remaining premium.
Adding insult to collapse, ETHZilla pivoted. Its subsidiary, ETHZilla Aerospace, now seeks to provide tokenized exposure to leased jet engines. Another subsidiary purchased 95 manufactured and modular home loans for $4.7 million to tokenize on Ethereum Layer 2. A company that raised $565 million to be an Ethereum treasury is now tokenizing aircraft engines. The market has priced this pivot accordingly.
The DAT model contains a reflexivity trap that works brilliantly in bull markets and catastrophically in bears.
The bull market loop: Rising crypto prices → stock premium to NAV expands → company issues cheap equity or debt → buys more crypto → prices rise further → premium expands further.
The bear market loop: Falling crypto prices → premium compresses to discount → capital markets access closes → company cannot buy more crypto → stock falls faster than the underlying asset → forced selling or dilution to meet obligations.
This is not a theoretical risk. It is happening now across the DAT sector. As Slate documented in February 2026, at least 168 public companies are playing some version of this game. The overwhelming majority have business models even thinner than Strategy's $500 million annual software revenue.
The core problem is that DATs are leveraged bets marketed as corporate strategies. Strategy carries $8.2 billion in convertible bonds and $8 billion in preferred shares, with an 11.25% dividend obligation on its STRK preferred stock. In a sustained bear market, the company must either continue diluting shareholders through ATM share sales, find new debt markets willing to lend against depreciated collateral, or — the scenario Saylor insists will never happen — sell Bitcoin.
As Anton Golub, founder of decentralized exchange Freedx, argued: "Strategy will dilute shareholders by issuing new shares, dump on retail via ATM sales, to raise cash to pay hedge funds."
MARA Holdings represents a particularly acute version of the DAT crisis: a Bitcoin miner whose market capitalization ($3.05 billion) has fallen below the value of its Bitcoin holdings ($3.69 billion).
On February 6, MARA moved 1,318 BTC worth approximately $86.9 million over a 10-hour period to multiple counterparties. The largest tranche — 653 BTC worth $43.4 million — went to digital asset manager Two Prime. Additional transfers of 200 BTC and 99.99 BTC went to BitGo-linked addresses, with another 355 BTC sent to unidentified wallets.
MARA's stock plunged 18.72% on the day of the transfers, closing at $6.73 — a 35% monthly decline. While the company characterized the moves as custody management rather than liquidation, the market's reaction was unambiguous. When a Bitcoin treasury company moves $87 million in BTC to counterparties during a market crash, investors assume the worst.
The miner-treasury hybrid faces a double squeeze: mining economics deteriorate as Bitcoin's price falls (reducing revenue per hash), while the treasury component suffers the same NAV compression afflicting all DATs. MARA's effective position is a leveraged long on Bitcoin's price with the operational costs of a mining business layered on top — a structure that amplifies losses in both directions.
In its February 16 filing, Strategy made an extraordinary claim: it can survive even if Bitcoin drops to $8,000 per coin.
The math is technically correct but misleading. At $8,000, Strategy's 714,644 BTC would be worth approximately $5.7 billion — roughly covering its $6 billion in net debt. The company would be solvent but effectively worthless as an equity investment. The $54 billion invested in Bitcoin would have produced a $48 billion loss.
More critically, the stress test ignores the path dependency. Strategy's convertible bonds begin maturing in 2027, with the earliest tranche covering approximately $1.2 billion. If Bitcoin remains well below the conversion prices at maturity, bondholders will demand cash repayment rather than converting to equity. Strategy's software business generates approximately $500 million in annual revenue — insufficient to service the debt without selling Bitcoin or raising new capital.
Michael Saylor's stated plan is to "equitize" the convertible debt over three to six years — converting bonds into equity to eliminate the maturity risk. But equitization requires the stock to trade at or above each bond's conversion price. With MSTR trading at a discount to NAV, that mechanism is currently broken.
The $8,000 stress test is a solvency argument, not a viability argument. Strategy can survive extreme scenarios without defaulting. But survival at $8,000 Bitcoin means years of dilutive share issuance, potential forced selling, and a stock price that may never recover to its 2024 highs.
Not all DATs will fail. The model separates into two categories:
Survivors: Companies with genuine operating businesses that use Bitcoin as a treasury asset alongside real revenue streams. Strategy, despite its challenges, has $500 million in software revenue and five years of conviction-tested holding. Companies in this category can weather extended bear markets because their equity isn't solely dependent on crypto prices.
Casualties: Pure-play copycats that raised capital solely to buy and hold crypto, with no operational business, no cost advantage, and no demonstrated willingness to hold through drawdowns. ETHZilla is the archetype. These companies offered investors nothing that a spot ETF couldn't provide more efficiently, and the market is now pricing that redundancy at near-zero.
The critical variable is conviction asymmetry. The DAT model requires the corporate buyer to hold when everyone else is selling. Any company that sells into weakness — as ETHZilla did — destroys the only value proposition that justifies its existence as a going concern.
The broader lesson is economic: financial intermediaries that exist solely to provide access to an asset lose their value when direct access becomes available. Bitcoin ETFs are the direct access vehicle. The DAT middleman era is ending.
The Digital Asset Treasury model was a brilliant financial innovation born from a specific market condition: institutional demand for Bitcoin exposure with no regulated vehicle to provide it. That condition no longer exists. Spot Bitcoin ETFs, approved in January 2024, offer everything a DAT offers — liquid, regulated Bitcoin exposure — without the corporate overhead, debt risk, management uncertainty, or premium-to-NAV dependency.
What remains is a $43 billion stress test. Michael Saylor's conviction is being tested against the most hostile market conditions the model has ever faced: Bitcoin below his average cost basis, a stock trading below NAV, $8.2 billion in convertible bonds with ticking maturity clocks, and an 11.25% preferred dividend obligation that demands cash regardless of market conditions.
The most likely outcome is bifurcation. Strategy survives through a combination of conviction, dilution, and eventual Bitcoin price recovery — scarred but intact. The copycat DATs, numbering in the hundreds, collapse into irrelevance as their stocks converge toward the liquidation value of their crypto holdings minus debt. The middleman premium was the product, and the market no longer needs the middleman.