For years, listed Bitcoin miners promoted a simple and powerful idea: mine BTC, hold it, and let the balance sheet compound alongside the asset. That “never sell” ethos helped miners position themselves as leveraged public-market proxies for Bitcoin itself. But that framework is changing. MARA Holdings, in its latest annual filing, disclosed that it has expanded its crypto management strategy for 2026 to permit sales of Bitcoin held on its balance sheet. This is a notable departure from the stricter policy it used in 2025, when sales were only allowed from newly mined production rather than from legacy treasury reserves.
The distinction matters. A miner that sells newly mined BTC is simply managing operating cash flow. A miner that allows sales from treasury inventory is reclassifying its reserve from something effectively untouchable into a strategic pool of capital. MARA did not say it would immediately liquidate its holdings, and the filing does not imply a forced exit from Bitcoin exposure. Instead, it creates a formal framework under which the company can buy or sell BTC depending on market conditions and capital allocation priorities. In practice, that gives management far more flexibility than the old reserve-first model.
This shift comes at a time when the mining industry is under increasing pressure. Miner revenues remain closely tied to Bitcoin’s market price, but major costs such as electricity, infrastructure, debt service, and expansion capital are either fixed or rising. In bull markets, a large BTC treasury can amplify upside and strengthen the equity story. In weaker markets, the same treasury can magnify balance-sheet volatility and constrain financial flexibility. For public miners, treasury strategy is no longer a side issue. It sits at the center of how they fund growth, manage risk, and explain their business to investors.
Why MARA is loosening its long-standing hold strategy
As of Dec. 31, 2025, MARA held 53,822 BTC. Based on a year-end spot price of $87,498, those holdings were worth approximately $4.7 billion. At the same time, the company recorded a $422.2 million decrease in the fair value of its Bitcoin holdings as prices fluctuated during the year. That combination captures the basic trade-off of treasury-heavy mining models: substantial upside during strong markets, but equally meaningful accounting and valuation pressure when Bitcoin becomes volatile.
MARA’s filing also makes clear that not all of its Bitcoin is simply sitting idle. Roughly 28% of the company’s BTC was deployed in lending, trading, or collateral arrangements. Specifically, about 9,377 BTC had been loaned to counterparties, while another 5,938 BTC was pledged against $350 million in outstanding credit facilities. That is a significant detail, because it shows that even before authorizing direct treasury sales, the company was already treating Bitcoin as an active balance-sheet asset that could support financing, yield generation, and broader capital management.
Those activities produced measurable income. According to the filing, MARA generated $32.1 million in interest income from its lending arrangements. For a miner, this can be an attractive middle ground: retain Bitcoin exposure while still earning incremental returns on treasury assets. However, that strategy also introduces its own complications, including counterparty risk, collateral management, and reduced flexibility during periods of stress. When Bitcoin prices move sharply, the distinction between liquid reserves, pledged collateral, and loaned assets becomes far more important.
Operationally, MARA remains one of the largest mining companies in the sector. It operates roughly 490,000 mining rigs and reported 66.4 exahashes per second of energized hashrate at the end of 2025. Its total energy capacity stood near 1.9 gigawatts, and purchased energy costs reached $179.0 million during the year. These figures highlight the capital-intensive nature of industrial-scale mining. Even large, well-capitalized miners cannot rely solely on Bitcoin appreciation; they must continuously manage energy procurement, infrastructure deployment, and financing conditions.
MARA mined 8,799 BTC in 2025, down from 9,430 BTC in 2024. The company attributed that decline to the effects of the April 2024 halving and increasing network difficulty. This is one of the most important structural forces shaping miner behavior today. After the halving, block rewards are reduced, while competition across the network continues to intensify as other operators add capacity and improve machine efficiency. The result is that each coin becomes harder and more expensive to produce, making passive treasury accumulation less straightforward than it once seemed.
At the same time, MARA is moving beyond pure mining. The company is advancing plans to develop data centers tailored for artificial intelligence and high-performance computing workloads. Management has described its power-rich sites as well suited to customers that need reliable access to energy at scale. These projects can be attractive, but they also require significant capital and long planning cycles. In that context, treasury monetization becomes more than a view on Bitcoin prices. It becomes a practical source of funding for a business model that is growing more diversified and more infrastructure-heavy.
Core Scientific is making a similar treasury pivot
MARA is not alone. Core Scientific said this week that it expects to monetize substantially all of its Bitcoin holdings in 2026 as part of a broader transition toward AI and high-density colocation services. That statement suggests the change is industry-wide rather than company-specific. For some public miners, the next phase of competition may not be defined only by hashrate growth. It may depend just as much on how effectively they can repurpose their power assets, sites, and data center infrastructure for adjacent computing markets.
The company has already started acting on that strategy. In January, Core Scientific sold about 1,900 BTC for approximately $175 million, implying an average sale price near $92,000 per coin. At the end of 2025, it still held 2,537 BTC worth about $222 million. Those figures indicate that the company is not dumping Bitcoin in distress. Instead, it appears to be monetizing treasury assets in a controlled way to support a broader operational transition.
Core Scientific has also said that its mining segment is being maintained primarily to satisfy power commitments while its sites are converted into facilities designed to support AI and other compute-intensive workloads. That is an important nuance. Mining is still part of the business, but increasingly as a bridge function rather than the sole long-term focus. In effect, the company is using its existing mining infrastructure to preserve utilization and revenue while it repositions those assets for higher-value applications.
Financially, Core Scientific ended 2025 with about $530 million in liquidity. It also outlined multibillion-dollar financing potential connected to data center contracts. In a higher-rate environment, selling BTC can be more attractive than issuing additional equity or taking on more debt. It gives the company immediate cash, reduces financing dependence, and can improve execution certainty for large infrastructure projects. That makes treasury sales a strategic capital allocation decision rather than simply a reaction to market weakness.
The trade-off, of course, is reduced direct exposure to Bitcoin’s upside. Historically, one of the main attractions of public miners has been their sensitivity to BTC appreciation. Investors often viewed them as amplified Bitcoin vehicles: if BTC rose, miners’ inventories, margins, and equity narratives could all strengthen at once. But if miners increasingly choose to sell treasury reserves and emphasize AI, colocation, and digital infrastructure services, investors may begin to value them less as Bitcoin proxies and more as hybrid infrastructure companies.
Post-halving economics are making treasury policy central
The larger lesson is that the mining sector has entered a more complex operating phase. After the halving, miner economics tightened, and network competition continued to rise. That means management teams now have to make difficult choices not only about machine deployment and energy sourcing, but also about treasury structure. Should BTC be held for upside, lent for yield, pledged as collateral, or sold to fund expansion? These decisions are now as consequential as decisions about site development, power procurement, and entry into adjacent compute markets.
Each treasury option comes with its own advantages and costs. Holding preserves maximum Bitcoin upside. Lending can generate interest income, as MARA’s $32.1 million result shows. Pledging BTC can unlock credit capacity, as seen in MARA’s use of 5,938 BTC against $350 million in facilities. Selling, meanwhile, is the most direct way to strengthen cash balances and fund near-term growth. But every path also involves sacrifices: holding means tolerating volatility, lending introduces counterparty exposure, pledging creates collateral constraints, and selling gives up some future appreciation.
What makes the current moment particularly significant is that treasury policy is now intertwined with strategic reinvention. Both MARA and Core Scientific are linking BTC monetization, directly or indirectly, to opportunities in AI and high-performance computing. They are not merely adjusting financial policy in isolation. They are responding to a structural shift in which power-rich mining sites may be redeployed into broader compute infrastructure businesses. The value of a miner increasingly lies not only in the coins it holds or the hashrate it controls, but in how effectively it can convert energy access and physical infrastructure into durable cash flow.
At the time of writing, BTC is trading below $67,000 after briefly topping $70,000 yesterday. The current price is $66,741.91. That price action itself helps explain why miners are rethinking treasury practices. When Bitcoin pulls back, the effects are not limited to mark-to-market values. Liquidity planning, financing choices, expansion schedules, and investor expectations all come under pressure. For a company simultaneously building AI-ready data centers or reshaping its business mix, turning some BTC into cash can be more practical than maintaining a strict ideological commitment to never sell.
That does not necessarily mean these miners have become bearish on Bitcoin. A more plausible interpretation is that the sector is maturing. As mining companies grow larger, carry heavier capital expenditure requirements, and diversify into infrastructure and compute services, Bitcoin becomes one part of a broader financial and operational system rather than a sacred reserve that can never be touched. MARA and Core Scientific are showing the market that in the post-halving era, success may depend as much on treasury discipline and capital flexibility as on raw mining scale.