MARA Holdings could be ready to test the current BTC financial meta. Major miners have accumulated BTC as strategic funds rather than treating it as working capital. Changes can have an impact far beyond a single company.
The company’s March 2 filing authorizes the balance sheet sale of its entire treasury of 53,822 BTC, representing a complete reversal of its policy of “holding all mined and purchased Bitcoin for the time being” in 2024.
Bitcoin is trading at around $68,000, down nearly 46% from its highs in late 2025, while the market has thinned to the point where modest selling can have a big impact.
The timing raises the question of what would happen if one of the industry’s largest holders treated Bitcoin as working capital rather than a matter of faith.
A policy that was never supposed to change
MARA’s 2024 10-K positions MARA alongside Strategy as a Bitcoin maximalist.
The pivot began in late 2025, with MARA selling approximately 4,076 BTC for $413.1 million, with an implied average of $101,000 per BTC. The 2026 filing would allow the balance sheet sale, making Bitcoin a “readily convertible source of liquidity.”
Three factors increase risk.
First, 15,315 BTC, or 28% of its holdings, has been pledged as a loan or collateral. This leaves 38,507 BTC unrestricted, with an issuance of $2.6 billion or 60 days after the halving.
Second, MARA recorded a fair value decline of $422.2 million and a trading loss of $69.1 million in 2025.
Third, MARA partnered with Starwood Capital to develop AI data centers targeting 1 GW over 2.5 GW, capital-intensive infrastructure to unlock liquidity needs.
Logic: Realize fund management and AI by selling BTC instead of diluting shareholders. This trade-off transforms MARA from a Bitcoin ETF to a capital allocation holding a volatile asset.
timing is not random
Three drivers gathered together to discuss “Why now?”
First is balance sheet pressure. After the halving, the reward was reduced to 3.125 BTC, but the difficulty and energy cost squeezed profits.
Despite the hashrate increasing to 66.4 EH/s, production decreased by 7% to 8,799 BTC. Liquidity becomes an emergency as Bitcoin falls from the $76,000-$126,000 range to $60,000.
The company has $350 million in convertible debt due in 2027.
The second is capital investment for AI. MARA’s Starwood partnership targets sites that switch between Bitcoin mining and AI computing. Starwood will lead design and construction. MARA contributes to the site and retains up to 50% ownership.
This is a bet that the monetization of computing power will outweigh the mining revenue after the halving.
Third, the microstructure of the market. Liquidity has been deteriorating since late 2025, with spot trading volumes 25-30% below last year’s levels. As a discretionary seller, MARA does not need to crash the market. Instead, when emotions are vulnerable, it creates an overhanging narrative.
MARA officially decided to do this not in spite of the weak conditions, but because the weak conditions made selling BTC more reliable compared to more expensive funding.
MARA is not the only overhang
Public miners held a total of 116,697 BTC, down 4.42% month-on-month.
MARA’s 53,822 BTC is almost half of the total. The broader pool includes Riot Platforms (18,005 BTC), CleanSpark (13,513 BTC), Hut 8 (10,278 BTC), and Core Scientific (2,537 BTC).
Core Scientific expects to monetize “substantially all” of its holdings in 2026. In January, he sold 1,900 BTC for $175 million, or $92,000 per coin. Bitdeer liquidated its entire finances in late February.
If the economics of AI infrastructure trump hashrate expansion, miners are now treating Bitcoin as inventory to monetize.
The question is how quickly and at what scale will other companies follow suit, and we frame that in three scenarios.
In a conservative scenario, miners sell their production, but the treasury keeps it. A 10% drawdown outside of MARA is equivalent to 6,287 BTC, or 14 days of issuance.
In moderate cases, miners will fund AI capital investment by selling 5% to 10% of their holdings. For MARA, this corresponds to 2,700 BTC to 5,400 BTC, or 6 to 12 days from issuance. This equates to between $180 million and $361 million.
A 25% bulk drawdown will free up 29,174 BTC, or 65 days of issuance.
In the aggressive scenario, a 50% drawdown would bring 58,349 BTC to the market, equivalent to 130 days of new supply. Risk is about story, not quantity.
Bitcoin’s 24-hour trading volume exceeds $50 billion, but when multiple miners become known as sellers during macro stress, the impact extends beyond spot prints to sentiment and derivatives positioning.
MARA applications allow others to comply without showing signs of distress.
| scenario | who sells | BTC amount | EST (Eastern Standard Time). Notional amount (approximately $68,000) | Equivalent to “new issuance days” (~450 BTC/day) |
|---|---|---|---|---|
| conservative | Non-MARA miners (10% drawdown) | 6,287BTC | ~$428 million | ~14th |
| Moderate (MARA) | MARA sells 5-10% of its holdings | 2,700–5,400 BTC | ～$184 million～$367 million | Approximately 6-12 days |
| Medium (industry) | Public Miners Collective (25% drawdown) | 29,174BTC | ~$2 billion | ~65 days |
| aggressive | Public Miners Collective (50% drawdown) | 58,349BTC | ~$4 billion | ~130 days |
What change reveals
In addition to the three scenarios, three competing narratives emerge.
The first is the AI pivot. Miners repurpose power infrastructure into data centers and use Bitcoin as a funding source.
MARA’s Starwood partnership targets AI-enabled infrastructure with toggle economics. This is a strategic reallocation from power certainty to capacity certainty.
The second story is tactical risk management. After a fair value decline of $422.2 million and a trading loss of $69.1 million, MARA treats Bitcoin as a management position.
Its shallow depth and high macro sensitivity make it a valuable discretionary liquidity tool.
The final story is a structural regime change: the end of minor HODL. The contrast between “holding all BTC” in 2024 and “may buy and sell from time to time” in 2026 shows that miners are acting like capital allocators, optimizing returns across mining, grid services, and AI leasing.
Each story contains different supply implications.
When an AI pivot occurs, BTC sales funds will be transferred. In this case the supply pressure is front loaded but finite.
If the risk management story is moving forward, sales will track volatility and miners will be countercyclical sellers.
Finally, the regime shift means that approximately 117,000 BTC of miner funds will be subject to active management, changing baseline assumptions regarding supply absorption.
important watch
The next thing to become clear is MARA’s Q1 10-Q form, expected in mid-May.
Investors will be scrutinizing how much BTC has been monetized since the policy change, whether the AI milestone is tied to Treasury withdrawals, and what guidance is provided on minimum reserves and pace of selling.
The gap until May creates a narrative void that the macro situation fills.
Bitcoin trades in a risk-off manner due to energy shocks and inflation concerns, precisely when “those who may be forced to sell” are in the ascendant.
MARA’s filing does not state that it will sell a majority stake. Still, if liquidity is thin enough and execution method determines whether a $1 billion sale is quietly absorbed or extends the downside, approval alone creates a price-sensitive reference.
Starwood’s timeline adds urgency. The partnership is targeting 1 GW in the short term and up to 2.5 GW in the future, although “short term” is undefined.
As MARA accelerates construction to capture AI demand, funding needs will be compressed. If growth is slow, BTC sales can span several years. That will determine whether MARA’s finances will be a multi-year drag or a one-time recapitalization.
If Q1 earnings reveal that multiple miners are expanding sales licenses or tying BTC monetization to AI capex, the market will reprice the entire miner financial base as supply overhang rather than strategic reserves.
This repricing does not require actual sales, it simply means that investors will no longer treat their miner holdings as a fixed supply.
what’s actually at stake
MARA changes are less about what you allow and more about what you signal.
For four years, miners have positioned government bonds as a differentiator by aligning stock performance with BTC price appreciation. It worked when Bitcoin was rising, capital was cheap, and the post-halving economy was theoretical.
Bitcoin is currently down nearly 50% from its highs, capital markets favor AI over cryptocurrencies, and post-halving margins are tighter than the model.
If MARA successfully executes the AI pivot and uses the BTC sales as one-time funding, the Treasury reversal story will end cleanly. If AI projects drag on or Bitcoin recovers faster than expected, miners could sell reserves at cyclical lows to fund struggling projects.
For the cryptocurrency market, the stakes are clear.
Miner treasuries were one of the last bastions of non-speculative Bitcoin demand and represented entities that accumulated Bitcoin for operational purposes.
If that group moves to active management, Bitcoin will lose structural bids and gain structural sellers. When the world’s largest Bitcoin miner by holding publicly announces its ability to sell its entire stack, it signals that even its followers are hedging.
(Tag translation) Bitcoin
