I still remember the first time I saw a major mining company openly talk about selling large amounts of Bitcoin instead of holding everything. It felt strange. For years the narrative was simple: miners mine, they hold, and the supply gets tighter. Then 2026 arrived and the numbers started telling a very different story. Fourteen publicly listed companies collectively sold 108247 Bitcoin, a figure that sits at roughly 8.08 billion dollars depending on the price levels used in the tally. That is not a small adjustment. That is a genuine shift in how some of the biggest names in the industry manage their balance sheets.
Why Public Companies Started Selling So Much Bitcoin
The scale alone forces you to pause. More than one hundred thousand coins left corporate wallets in a single year. Most of the volume came from companies that actually produce Bitcoin every day. Miners. The very firms that once treated every coin as a long-term reserve suddenly treated large portions of their inventory as working capital.
Lookonchain’s November tally put the total at 108247 Bitcoin. MARA Holdings and Riot Platforms alone accounted for more than half of that figure. Together they moved 60845 coins. When two names control over 56 percent of the recorded sales, you know the rest of the list is interesting but secondary. Still, the secondary names matter because they reveal a broader pattern rather than an isolated decision by two aggressive operators.
MARA’s Large Scale Liquidations and Debt Clearance
MARA sold approximately 26984 Bitcoin for about 1.9 billion dollars in the first nine months of the year. After the quarter ended the company sold another 13597 coins for roughly 1.1 billion. That second wave was not random. A large part of the cash went straight toward paying off Bitcoin-backed loans. The company cleared 750 million dollars of secured debt and, in the process, released 17477 Bitcoin that had been sitting as collateral.
I find this particular detail revealing. Paying down debt that was itself backed by Bitcoin creates a circular kind of freedom. Once the loans disappear, those previously locked coins become available again. MARA also updated its treasury policy in 2026 so that it could sell coins it already held, not only newly mined production. That policy change quietly removed an important psychological barrier. Holding forever is no longer the only acceptable stance.
Later in the year MARA priced notes related to its Long Ridge Energy acquisition. The company planned to use a mix of note proceeds and existing cash. Even while expanding into power and infrastructure, the firm continued treating Bitcoin as a flexible asset rather than an untouchable reserve. That dual approach feels pragmatic more than ideological, and pragmatism tends to spread once it proves useful.
Riot Platforms Sold More Than It Mined
Riot’s numbers are even more striking when you set production against sales. The company recorded 1.484 billion dollars from selling 20264 Bitcoin through September. During the same period it mined only 4697 coins. Holdings dropped from 18005 Bitcoin at the start of January to just 2418 by the end of September. That is a dramatic reduction.
The cash went toward data center development and the full repayment of a 200 million dollar credit facility. Riot has been clear that it expects to keep selling substantially all newly mined Bitcoin to cover operating costs, development spending, and working capital. In other words, the company has moved into a model where production is monetized almost immediately rather than accumulated.
Third-quarter mining costs offer useful context. Before equipment depreciation the cost sat at 39265 dollars per coin. With depreciation included the figure rose to 78805 dollars. The company valued the Bitcoin it produced during the quarter at an average of 70855 dollars. Power curtailment credits added another 24.7 million dollars of value. Those credits matter. They show how miners can generate revenue by reducing electricity use under certain grid conditions, adding another layer to the overall economics.
Core Scientific and the AI Hosting Pivot
Core Scientific sold 3374 Bitcoin for 286.8 million dollars through September while mining only about 869 coins. Holdings fell from 2537 Bitcoin at the end of 2025 to just 32 coins by the end of the third quarter. Almost everything was liquidated.
The company directed the proceeds toward planned capital spending and other cash needs. Much of that spending supports data center build-outs under AI hosting agreements. CoreWeave funded 180.9 million dollars of Core Scientific’s capital expenditures in the first nine months. The shift from pure Bitcoin mining toward high-performance computing hosting is no longer theoretical. It is visible in the balance sheet and in the near-zero Bitcoin inventory.
When a miner sells nearly four times what it produces in a given period, the message is hard to miss. Bitcoin is being treated as a funding source for the next chapter of the business rather than the primary product that must be held at all costs.
Keel Infrastructure’s Accelerated Exit from Mining
Keel Infrastructure, previously known as Bitfarms, exchanged 3169 Bitcoin for 239.6 million dollars in cash through September. Of that total, 2520 coins changed hands in the third quarter alone and generated 192.2 million dollars. By the end of September the company recorded only a nominal Bitcoin balance valued at roughly 36000 dollars.
Keel ended United States mining operations in June while preparing sites for high-performance computing. Legacy mining assets in Canada continue to operate, yet the overall direction is clear. Bitcoin sales funded the transition. The speed of the third-quarter sales suggests management wanted the cash on hand before pushing further into the new business model.
Cipher Digital and Steady Inventory Reduction
Cipher Digital sold 2252 Bitcoin for 168.9 million dollars through September. The company received 1095 coins from its own mining and another 34 coins from investee companies. Holdings dropped from 1433 Bitcoin at the beginning of the year to 310 coins by quarter-end. The reduction is less dramatic than some peers, yet it still shows a deliberate preference for cash over additional inventory.
These smaller but consistent sales across several names add up. When you stack them next to the outsized moves by MARA and Riot, the aggregate number becomes easier to understand.
Other Notable Contributors to the Tally
Lookonchain ranked Bitdeer Technologies third with an estimated 10553 Bitcoin in outflows between January and September. Cango appeared with 7001 coins and CleanSpark with 5715 coins. Of CleanSpark’s total, 3654 Bitcoin were delivered in connection with options or basis trades rather than simple open-market sales. Those structured deliveries still remove coins from the company’s effective exposure, so they belong in the overall count.
Sequans Communications, a semiconductor firm, announced the end of its Bitcoin treasury strategy and sold its final 314 coins after June 30. KULR Technology Group sold its entire 1099.09 Bitcoin balance for roughly 80.1 million dollars during the September quarter and used part of the proceeds to repay a 20 million dollar Coinbase loan. These non-miner exits reinforce the idea that corporate Bitcoin strategies are being reassessed across different sectors, not only among pure-play miners.
What the Numbers Actually Mean for Market Supply
One hundred eight thousand Bitcoin is a meaningful volume. In any given year the network produces a fixed amount of new coins through the block reward. When public companies release more than one hundred thousand coins that were previously sitting in corporate treasuries or were newly mined and quickly sold, the net effect is additional supply hitting the market.
I have watched enough cycles to know that concentrated selling from a small group of well-known names can create temporary pressure even if the broader ownership base remains strong. The difference this time is that the sellers are not random holders or leveraged traders. They are the companies that control a large share of hash rate and that file detailed quarterly reports. Their decisions are visible, measurable, and often explained in plain language inside those filings.
Perhaps the most interesting aspect is how openly these firms now discuss the trade-off between holding Bitcoin and funding growth. Data center construction, AI hosting contracts, debt repayment, and working capital needs have become legitimate reasons to sell. The old pure HODL stance has softened into a more flexible corporate finance approach.
The Shift Toward High-Performance Computing
Several of the companies on the list are simultaneously building or converting facilities for AI and high-performance computing workloads. That transition requires capital. Bitcoin sales have become one of the faster ways to raise that capital without issuing large amounts of new equity or taking on expensive traditional debt.
In my view this dual-use strategy makes sense for operators that already control power contracts and large land parcels. The same infrastructure that supports mining can, with modifications, support denser computing loads. Selling Bitcoin to fund that pivot is simply using one asset to build another. Whether the long-term returns from AI hosting exceed the opportunity cost of the sold Bitcoin remains an open question, but management teams appear willing to make the bet.
Treasury Policy Changes and the End of Automatic Accumulation
MARA’s decision to expand its treasury policy so that existing holdings could be sold is more important than it might seem at first glance. For years many mining companies operated under informal or formal rules that treated newly mined coins as automatic additions to the treasury. Changing that rule removes a structural source of continuous accumulation.
Once a company grants itself permission to sell both production and existing inventory, the decision becomes purely financial. Price levels, capital needs, debt covenants, and growth opportunities all enter the calculation. That is healthy corporate behavior, yet it also means the market can no longer assume that miner production will stay locked away indefinitely.
When the biggest producers start treating Bitcoin as working capital rather than permanent reserves, the supply dynamics change in ways that simple production numbers no longer capture.
How Investors Might Interpret the Selling Wave
Some observers will view the sales as a sign of weakness or lack of conviction. Others will see disciplined capital allocation. Both interpretations contain partial truth. Selling to repay high-cost debt or to fund contracted AI hosting revenue can improve the overall risk profile of a company. Selling simply because the price looks attractive in the short term is a different decision.
The filings generally emphasize the former set of reasons. Debt clearance, data center development, and working capital appear repeatedly. That consistency across multiple companies suggests a sector-wide recalibration rather than isolated opportunistic trades.
For long-term holders outside these firms, the practical effect is that more coins have entered circulation. Whether those coins stay in strong hands or move into weaker ones depends on who the ultimate buyers were. Public data on the destination of each sale is limited, so the secondary market impact remains partly opaque.
Looking at the Broader Corporate Landscape
Not every public company that holds Bitcoin is selling. Some continue to accumulate or simply maintain existing positions. The 108247 figure therefore represents a specific subset of firms that chose to reduce exposure. The existence of that subset is still notable because it includes several of the most visible and actively followed names in the mining sector.
I have found that markets often overreact to visible selling while under-reacting to quiet accumulation. The opposite can also occur. The key is to separate the mechanical effect of additional supply from the narrative that surrounds it. Additional supply is real. The narrative that “everyone is selling” is usually exaggerated.
Fourteen companies is not the entire industry. Yet when those fourteen include the largest pure-play miners by hash rate and market capitalization, their collective behavior carries weight.
Cost Structures and the Incentive to Sell
Riot’s disclosed mining costs illustrate why selling can look rational even at prices that once seemed high. A fully loaded cost near 79000 dollars creates a different decision matrix than a cash cost below 40000 dollars. When depreciation is included, the economic picture becomes more conservative. Companies that report both figures give investors a clearer view of the real pressure points.
Power curtailment credits add another variable. Earning meaningful revenue by reducing load during peak grid periods improves the overall economics of a mining operation. Those credits do not remove the incentive to sell production when capital needs arise, but they do soften the impact of higher electricity costs in certain markets.
The Role of Structured Trades and Options Deliveries
CleanSpark’s inclusion of 3654 Bitcoin delivered through options or basis trades is worth highlighting. Not every outflow is a simple market sale. Some companies use derivatives to manage price exposure or to generate additional yield. When the contracts settle, the coins leave the treasury even if the firm never placed a traditional sell order on an exchange.
Those structured deliveries still reduce the company’s net Bitcoin exposure. From a supply perspective they function much like ordinary sales. Investors who track only open-market volume can miss this portion of the total.
What Comes Next for Corporate Bitcoin Strategies
The 2026 sales wave does not automatically mean that public companies will continue liquidating at the same pace. Capital needs change. Debt levels fluctuate. AI hosting contracts either generate the expected cash flow or they do not. Price levels themselves influence the decision to sell or hold.
What has changed is the willingness to treat Bitcoin as a flexible corporate asset. Once that mental shift occurs, future decisions become more situational. In periods of high capital expenditure or elevated interest rates, sales become more likely. In periods of strong free cash flow and lower external funding needs, accumulation can return.
I suspect we will see more companies adopt explicit dual policies that allow both accumulation and opportunistic sales depending on internal capital requirements. The pure “never sell” stance that dominated earlier years is giving way to something more nuanced.
Practical Takeaways for Market Participants
Anyone tracking Bitcoin supply dynamics should now include a regular review of public company filings and treasury updates. The old assumption that miner production stays locked is no longer reliable. Quarterly reports from the largest operators have become essential reading.
- Monitor the largest miners’ production versus sales figures each quarter
- Watch for changes in stated treasury policies
- Pay attention to debt repayment announcements that involve Bitcoin collateral
- Track capital expenditure plans tied to AI or high-performance computing
- Note any structured derivative deliveries that reduce effective holdings
These simple habits provide a clearer picture of near-term supply pressure than network production numbers alone.
Balancing Growth Ambitions with Bitcoin Exposure
The tension between growing the business and maintaining Bitcoin exposure is real. Companies that want to expand into energy, data centers, or AI infrastructure often need cash faster than mining operations alone can provide. Selling Bitcoin accelerates that cash generation. Holding Bitcoin preserves upside if prices rise significantly. Both choices carry opportunity costs.
Management teams are paid to navigate those trade-offs. Shareholders will ultimately judge the results. For now the visible pattern in 2026 is that a meaningful group of public companies chose cash and growth projects over maximum Bitcoin accumulation.
Whether that choice proves wise will depend on the returns generated by the new projects and on the future path of Bitcoin itself. The decision, however, has already been made and recorded in the filings.
Final Thoughts on a Changing Corporate Mindset
The sale of 108247 Bitcoin by fourteen public companies marks more than a statistical footnote. It reflects a broader evolution in how corporate treasuries view digital assets. Flexibility has replaced rigidity. Balance sheet management has taken priority over pure ideological holding.
That evolution is neither purely positive nor purely negative. It simply is. Markets adapt to the behavior of large participants. Investors who notice the change early and adjust their own assumptions about supply accordingly will be better prepared for whatever comes next.
The companies that sold did so for stated reasons that mostly center on debt reduction, infrastructure investment, and operational funding. Those reasons are understandable. The volume is large enough to matter. And the transparency of the filings means the rest of the market can at least see the moves as they happen.
In the end, Bitcoin remains a scarce asset with a fixed issuance schedule. Corporate selling adds temporary supply. Corporate accumulation removes it. Both forces will continue to interact. The difference in 2026 is that the selling side of that equation became more visible and more concentrated than many expected at the start of the year.
Watching how these same companies behave in the coming quarters will tell us whether the wave was a one-time recalibration or the beginning of a more permanent shift in corporate strategy. For now the data is clear. A significant amount of Bitcoin left public company treasuries, and the market absorbed it. What happens next is the part still being written.