MARA is repositioning from a pure Bitcoin miner to a capital-efficient AI and data-center infrastructure player, leveraging a deleveraging strategy to potentially support a higher valuation amid Bitcoin volatility.

The company is restructuring its capital by liquidating part of its Bitcoin treasury to clean up the balance sheet, which could alter relationships with lenders, partners, and customers as it expands AI infrastructure and data centers.

The liquidation aims to decouple MARA’s valuation from Bitcoin swings, positioning the firm to pursue AI-driven revenue through higher-margin, long-term contracts and potential multi-year colocation deals with Big Tech and sovereign governments.

Industry context shows profitability pressures in the sector due to halving effects, high energy costs, and mining expenses, prompting a reassessment of treasury strategies and diversification.

Analysts maintain a moderate buy on MARA with a mean price target near $16, implying significant upside from current levels.

Bearish scenarios forecast a path to roughly $1.3 billion in revenue and $162 million in earnings in 2028, highlighting risks from crypto regulation and reliance on AI data-center initiatives.

Risks include tighter balance-sheet optionality if projects underperform or ramp slowly, and execution risk around site conversions, tenant demand, and power contracts.

Despite optimism for AI-driven demand, some experts warn that AI pivots don’t guarantee success, given scrutiny and the unproven scalability of AI compute for many hyperscalers.

Reactions to the move are mixed, with some praising treasury management and others criticizing the strategy.

The sale reflects a broader trend of crypto miners rethinking capital strategies to manage debt, fund operations, or invest in new lines like AI.

J. Wood Capital Advisors served as financial advisor and Paul, Weiss as legal counsel for MARA’s transactions.

Industry-wide shift toward AI is evident, with peers like Core Scientific, IREN, and CleanSpark pursuing AI cloud deals and AI-focused data-center expansion.