The companies that borrowed billions to buy Bitcoin are now selling it. At least seven publicly listed digital asset treasury firms have unloaded their Bitcoin holdings in recent weeks — some to repay debt, some to fund operations, and some to pivot entirely into artificial intelligence. The model that was supposed to revolutionize corporate finance is unraveling in real time.
By David Chen | July 24, 2026
The Hook: The DAT Model Breaks Down
It started as one of the boldest experiments in corporate finance. Strategy — formerly MicroStrategy — pioneered the digital asset treasury model in 2020, using its own cash reserves and borrowed funds to accumulate Bitcoin. The idea was simple: if you believe Bitcoin will appreciate faster than your debt costs, then borrow cheaply, buy Bitcoin, and let the spread work in your favor.
It worked spectacularly — for a while. A wave of imitators followed, including Satsuma Technology, Smarter Web Company, Sequans Communications, Nakamoto, and Empery Digital. These firms splurged cash and borrowed heavily to buy Bitcoin as the price climbed toward its record of roughly 126,000 dollars in October 2025.
Then Bitcoin fell about 50 percent from that peak. Share prices collapsed. And now, according to Matthew Sigel, Head of Digital Assets Research at VanEck, several of these companies have exited crypto entirely or are substantially reducing their holdings.
On-Chain Evidence: Who Is Selling and Why
The list of sellers reads like a who’s who of the treasury boom. Satsuma Technology shareholders just approved the liquidation of all 668 BTC, a return of capital, and a delisting from the London Stock Exchange. Another LSE-listed firm, Smarter Web Company, sold 178 BTC to repay a convertible instrument.
Smarter Web CEO Andrew Webley was candid about the reversal: while the company still recognizes the potential benefits of Bitcoin-denominated instruments, they no longer believe convertibles represent the right capital solution at this time.
- Satsuma Technology (SATS) — liquidating all 668 BTC, delisting from LSE
- Smarter Web Company (SWC) — sold 178 BTC to repay convertible debt
- Sequans Communications (SQNS) — sold 1,025 BTC, then disposed of nearly 80 percent of remaining holdings; 658 BTC left
- Nakamoto (NAKI) — sold roughly 284 BTC for working capital; shares down 99 percent from SPAC deal
- Strategy (MSTR) — sold approximately 3,620 BTC recently, authorized further sales for cash reserves
- Empery Digital — sold almost half its Bitcoin for buybacks and debt repayment
The Core Conflict: Miner Pivots and Management Shakeup
It is not just treasury companies unwinding. Bitcoin miners are also selling — but for a different reason. MARA Holdings and Bitdeer, two of the largest publicly traded mining firms, have been unloading Bitcoin to repay debt and repurpose their energy infrastructure for AI data centers.
The logic is brutal but clear. Mining Bitcoin at current prices — around 65,000 dollars, down from 126,000 dollars — may no longer cover operating costs at many facilities. Meanwhile, AI companies are desperate for computing power and willing to pay premium rates for access to energy-hungry data centers. For miners sitting on expensive infrastructure and cheap power contracts, pivoting to AI is the rational economic play.
The disruption goes beyond asset sales. Jack Mallers stepped down as CEO of Twenty One Capital, the Bitcoin treasury firm backed by Tether — a sign that even well-capitalized ventures are hitting turbulence. Meanwhile, Adam Back’s Bitcoin Standard Treasury Company (BSTR) failed to complete its proposed merger, scrapping the original SPAC terms due to unfavorable conditions.
Market Implications: Strategy’s New Math
Through all of this, Strategy remains the giant in the room. The company still holds more than 843,000 BTC — by far the largest publicly listed position. But even Strategy is adapting. The company just introduced a completely new metrics framework to give common shareholders a clearer picture of their actual exposure.
Under the old reporting, Strategy touted its gross Bitcoin holdings. The new framework introduces what the company calls Net Reserve: take the total Bitcoin and cash reserves of 58.8 billion dollars, subtract 6.8 billion dollars in out-of-the-money convertible debt and 15.5 billion dollars in preferred stock obligations, and you get a Net Reserve of 36.6 billion dollars.
That is a big haircut. It means roughly 22.3 billion dollars in senior claims rank ahead of common shareholders in any liquidation. The new BTC Breakeven ARR metric sits at 3.22 percent — meaning Bitcoin only needs to appreciate faster than that annual rate for Strategy to cover all its interest and dividend obligations indefinitely. At current prices, that bar seems achievable. But it is a striking shift from the unbridled optimism of 2024 and 2025.
The Verdict: Consolidation, Not Collapse
The digital asset treasury model is not dying — it is consolidating. The firms that borrowed aggressively at the top are being forced to face the music, and the weaker hands are folding. But Strategy, with its enormous stash and increasingly transparent reporting, is positioning itself as the survivor. CEO Michael Saylor even told investors the company may sell some Bitcoin to fund a dividend, saying it would inoculate the market against broader fears.
For investors watching from the sidelines, the lesson is straightforward. The treasury model works when Bitcoin goes up. When it goes down — especially when leverage is involved — the unwind can be painful and fast. Companies that over-leveraged at the top are learning this the hard way. Companies that managed risk and kept debt manageable will come out stronger.
The bigger story may be the miner-to-AI pivot. When the companies that physically secure the Bitcoin network start pointing their hardware at AI workloads instead, it tells you where the real money is flowing right now. Bitcoin mining is not going away — but the economics that defined the last cycle are shifting under everyone’s feet.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
Disclaimer: This article is for informational purposes only and does not constitute financial advice. Cryptocurrency investments carry risk; always do your own research.
satsuma liquidating all 668 btc and delisting lol. this is what happens when you copy microstrategy without the balance sheet to survive a 50pct drawdown
Smarter Web sold 178 btc just to repay one convertible. Webley basically admitted the whole thing was a failed experiment
the ones pivoting to AI are even worse. at least sell the btc and return capital instead of burning it on another buzzword
Stefan D. pivoting to AI is worse than just returning capital. at least liquidation is honest. burning BTC reserves on GPU clusters you dont understand is next level
called this a year ago. borrowing to buy a depreciating asset then acting shocked when the debt comes due. Satsuma liquidating 668 BTC is gonna hurt
saylor printed the playbook and nobody read the risk management chapter
ai_grift_ the AI pivot is the funniest part. you failed at treasury management so now you are pivoting to the other buzzword. shareholder money is play money to these people
The Smarter Web CEO basically admitted convertibles were a mistake. These companies were never Bitcoin believers, they were yield-chasing with shareholder money.
VanEck Sigel called the top on this trend months ago. 7 firms unwinding and counting, the copycat era is over
Strategy pioneered it and now 7 copycats are unwinding. the DAT model was always just leverage dressed up as innovation. BTC goes up, everyone is a genius. BTC goes sideways and the debt service eats you alive
speadsheet_og_ the worst part is the ones selling BTC to repay debt are creating sell pressure that hurts the remaining holders. vicious cycle
Smarter Web sold 178 BTC to repay one convertible note. the entire DAT model was a leveraged bet that BTC only goes up. it doesnt
pivoting to AI is the new pivoting to blockchain. same executives, same pitch deck template, different buzzword
saylor printed the playbook and these copycats skipped chapter 1 through 9. you need a real software business generating cash before you lever up on BTC
Satsuma liquidating all 668 BTC and delisting. copycats who skipped Saylors risk management chapters learning the hard way
Konrad B. Saylor survived because MSTR had a real software business generating cash. these copycats had nothing but BTC on the balance sheet
also saylor never sold a coin through three bear markets. these guys unwound at the first real drawdown, which is the exact opposite of the thesis they raised debt on
Saylor also never ran a premium-to-NAV model dependent on issuing equity. He just bought with cash flows. These copiers did leveraged mNAV arbitrage and called it the same playbook
saylor bought in 2020 before anyone could copy him. the copiers paid 2x NAV for BTC exposure with debt attached, that trade was broken at issuance
668 BTC gone in one move and then straight to delisting. no treasury strategy unwinds that fast, that was a bet hitting its stop loss
the tell was always the premium to nav. these things traded 2x nav on leveraged btc exposure while calling it a strategy. unwind was guaranteed the second btc flatlined for a quarter
the DAT math only works when your cost of debt is below bitcoins appreciation rate. literally no one modeled the year bitcoin goes sideways or down. seven firms unwinding in weeks is what that spreadsheet gap looks like in real time
^ exactly. and the converts had coercive redemption clauses nobody read. btc goes sideways for two quarters and suddenly your creditor can put the bonds back at par
strategy survived 2022 because no converts matured that year. the copiers stacked 2026 and 2027 maturity walls then acted shocked when the wall arrived on schedule
selling the bitcoin to pivot into AI is the funniest exit line of the cycle. the sophisticated treasury thesis ends with chasing whatever pumped last quarter