The 'never sell' era is over: Strategy is now selling Bitcoin. Here's what it does and doesn't mean

TL;DR

Strategy (formerly MicroStrategy), the company that built its identity on "never sell your Bitcoin," has now sold Bitcoin several times in 2026, some of it below its purchase price, disclosed in its own SEC filings. That is a genuine break from the past. But read it precisely: these are discretionary sales, authorised by the board in June, to fund fixed dividends on its preferred stock, not a forced liquidation, margin call or bankruptcy, all of which the company denies and its cash coverage argues against. The real stress is quieter: the stock now trades at a discount to the Bitcoin it holds, which breaks the money-machine that made the model work. Other treasury firms are cracking harder; several small ones have exited entirely.

For six years, one company defined the "Bitcoin treasury" trade: raise money on the stock market, pour it into Bitcoin, and never, ever sell. Michael Saylor turned "never sell your Bitcoin" into something close to a religious slogan. So when Strategy started actually selling Bitcoin in 2026, it looked like the whole thesis was breaking. The truth is more interesting than either the "Saylor capitulates" or the "nothing to see here" version. Here is what actually happened, what it means, and what it doesn't.

Did Strategy really sell Bitcoin? Yes

This part is not a rumour. Strategy has sold Bitcoin at least four times since late June 2026, each disclosed in its own regulatory filings, and some sales were made below the price it paid. The company still holds an enormous stack, about 840,447 BTC bought at an average of roughly $75,385 each, but after Bitcoin's fall to the low $60,000s (around $63,800), that hoard is worth some $10 billion less than it cost on paper.

The sales were authorised, not improvised. On 29 June 2026, Strategy's board approved a new "Digital Credit Capital Framework" that, for the first time, permits the company to sell Bitcoin, up to $1.25 billion of it, to cover about a year of obligations. Since then it has sold in tranches: roughly 3,588 BTC across late June and early July, then about 1,638 BTC in late July, then another 1,690 BTC in early August (raising about $109 million), all to service dividends on its preferred stock and to buy some of that preferred stock back. For a company whose entire brand was accumulation, this is a real regime change.

But it is not a forced sale, and the "never sell" line has a catch

Here is where the headlines overreach. Selling to fund a planned obligation is not the same as being forced to sell, and Strategy is emphatic on the distinction. Asked directly whether it was a distressed seller, CEO Phong Le told CNBC: "No, we're not a forced seller of bitcoin." The math supports him, at least for now. The company holds a multi-billion-dollar cash reserve against its roughly $1.76 billion a year in preferred dividends and interest, enough to cover those payments for over two years even before touching a single coin, and its nearest big debt maturity is not until February 2027. There is no reported margin call, no covenant breach, and no bankruptcy risk on the record. Anyone telling you Strategy is being "forced to liquidate" is inventing it.

Saylor himself has tried to square the slogan with the sales through a personal-versus-corporate distinction. "When I say 'Never Sell Your Bitcoin,' I speak as one saver to another," he said, adding that he has never sold his own coins, and that "Strategy is not my wallet." Whether you find that convincing or a dodge, the accurate description is narrow: the company has started selling to manage its finances, while its founder maintains the creed for individuals. Both can be true, and the "Saylor sold his Bitcoin" version circulating online is simply wrong.

The real crack: the premium has collapsed

The Bitcoin sales are the headline. The quieter story underneath is the one that actually matters, and it has a clunky name: mNAV, the ratio of a company's market value to the market value of its crypto.

For most of 2024 and 2025, Strategy's stock traded at a large premium to its Bitcoin, an mNAV well above 1 (it hit roughly 3.4x in late 2024). Investors were paying more than a dollar for every dollar of Bitcoin the company held, because they were also buying Saylor's ability to keep raising cheap money and buying more. In 2026 that premium evaporated. Strategy's basic mNAV has fallen to around 0.68x, meaning its shares now trade roughly 32% below the value of the Bitcoin sitting on its balance sheet. The market has stopped paying up for the machine, and started pricing the stock at a discount to its own coins.

That collapse is the deep problem, because the premium was not a vanity metric. It was the fuel.

How the machine works, and why running it in reverse hurts

The treasury-company model is what analysts call reflexive, a flywheel. When the stock trades at a premium to its Bitcoin, the company can issue new shares at that premium and use the proceeds to buy more Bitcoin. Because it sold shares dear and bought coins at fair value, each round increases the amount of Bitcoin backing every existing share, which justifies the premium, which lets it do it again. Rising Bitcoin plus a fat premium made it look like magic.

Run that in reverse and the magic curdles. Once the coin price falls and the premium disappears, issuing new shares no longer adds Bitcoin per share; it just dilutes existing holders. Analysts estimate the machine only creates value above roughly a 1.22x mNAV, so a company trading below that, let alone below 1, has effectively lost its cheap-financing engine. But the fixed bills, those preferred dividends and interest, do not pause. With cheap equity gone, a company can keep paying them only by issuing shares at a value-destroying discount or by selling coins, and over time, as any cash buffer is drawn down, that choice sharpens. Strategy, still sitting on years of dividend coverage in cash, has nonetheless chosen to let Bitcoin, rather than dilution, fund part of the bill early. That is the shift in one sentence: the flywheel stopped spinning, so the coins go from something you only ever buy to something you can also spend.

The short-seller Jim Chanos described the premium bluntly at its peak, calling terms like "Bitcoin yield" financial gibberish and the trade a simple arbitrage: "buying for $1, selling for $2.50." Tellingly, Chanos closed that bet in late 2025 as the premium collapsed, and is not currently short or predicting bankruptcy. The premium he was betting against is exactly the thing that has now gone.

It is not just Strategy

Zoom out and Strategy is the sturdiest example of a wider squeeze, not the worst.

MARA (the former Marathon) sold about 23,093 BTC in the first half of 2026, some $1.6 billion worth, to cut debt and fund operations, framing the move as de-levering on its own terms rather than distress.
Several smaller treasury companies have simply exited: firms like K Wave Media and Genius Group sold their last coins, some pivoting to AI instead, when the model stopped working for them.
The ETH-hoarding versions, BitMine (about 5.8 million ETH) and SharpLink, are trading at discounts to their Ethereum too, though both are still buying rather than selling.

And the mood is jumpy. On the very day this piece was written, a $320 million on-chain transfer from Japan's Metaplanet set off "they're dumping" rumours, until its CEO clarified it was "a routine custody operation. No bitcoin was sold." That is the environment: real stress makes every large wallet movement look like a fire sale, even when it isn't.

So are Bitcoin treasury companies cracking?

The honest answer splits in two. The model is genuinely under stress: the premium that powered it has vanished, the cheap-financing engine has stalled, and the weakest players have folded or sold out. That is real, and it is arguably the most important development in this corner of crypto all year.

But Strategy, the one everyone watches, is not collapsing. It is doing something more mundane and more telling: managing its obligations by selling, discretionarily, from a still-huge Bitcoin pile, while insisting it is not a forced seller, and, by the coverage math, being right about that for now. The risk is forward-looking, not realised: if Bitcoin stays low and the discount persists, the pressure to keep selling grows, and the February 2027 refinancing gets harder. The fair verdict is that the "never sell" era is over, the treasury flywheel has stopped, and the question has shifted from "how much Bitcoin can these companies buy" to "how gracefully can they carry what they already own." For more, see the Crypto section and our look at whether the 2026 bull run is over.

Strategy and the treasury trade, at a glance


Strategy's stack	~840,447 BTC at ~$75,385 avg; ~$10B underwater on paper
The change	Sold BTC 4+ times since late June 2026 (some below cost), authorised by a June board framework
Why it sold	To fund ~$1.76B/yr in preferred dividends and interest, not a margin call
Forced seller?	No: CEO denies it; 2+ years of dividend coverage in cash; nearest debt due Feb 2027
The real crack	mNAV ~0.68x, the stock trades ~32% below its own Bitcoin (was ~3.4x in 2024)
Others	MARA sold ~23,093 BTC; small firms exited; Metaplanet denied an Aug-13 sale rumour

Key takeaways

Strategy (formerly MicroStrategy) has genuinely sold Bitcoin several times in 2026, some below cost, disclosed in its own filings, ending its long "never sell" stance. It still holds about 840,447 BTC, roughly $10 billion underwater on paper.
These are discretionary sales authorised by a June 2026 board framework (up to $1.25B) to fund fixed preferred-stock dividends and interest (~$1.76B/yr), NOT a forced sale. The CEO says "we're not a forced seller," and the company has over two years of dividend coverage in cash and no debt due until February 2027.
The deeper stress is the collapse of Strategy's premium: its mNAV has fallen to about 0.68x, so the stock now trades ~32% below the value of its own Bitcoin, down from about 3.4x in 2024. That breaks the flywheel that let it issue shares at a premium to buy more coins.
With cheap equity financing gone but fixed bills remaining, the company must either dilute shareholders or sell Bitcoin, and it chose to sell. Short-seller Jim Chanos, who profited betting against that premium, closed the trade in late 2025 as it collapsed and is not predicting bankruptcy.
It is a wider squeeze, not just Strategy: MARA sold ~23,093 BTC in H1 2026, several small treasury firms exited entirely, and ETH-hoarders like BitMine trade at discounts too. But the biggest player is managing the strain, not cracking, so "the treasury era is over" overstates a real but non-catastrophic regime change.