•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Publicly traded Bitcoin mining company MARA Holdings (MARA) has substantially trimmed its corporate treasury, selling a total of 3,386 BTC during the first quarter of 2026.
The company is pivoting away from relying solely on cryptocurrency mining and moving toward high-performance computing (HPC) and artificial intelligence (AI) data centers. MARA said the proceeds from its BTC sales are being used to secure major acquisitions, including the Long Ridge Energy & Power compute campus.
The company’s stated goal is to apply its large energy resources to host AI workloads, positioning its infrastructure to support highly profitable computing demand.
MARA is also using its Bitcoin treasury to repair its balance sheet. The proceeds from BTC sales have been used to improve liquidity and fund the repurchase of convertible senior notes.
The BTC sell-off coincided with a difficult quarter. In Q1 2026, MARA reported a net loss of $1.3 billion. The loss was mainly attributed to a roughly 20% drop in Bitcoin’s price between January and March, which triggered a $1 billion impairment charge tied to the decline in crypto value.
Against this backdrop, selling portions of the treasury was necessary to shore up cash reserves.
Historically, miners have sold BTC to finance the purchase of more advanced ASIC mining rigs. MARA has indicated a shift away from that approach, confirming it does not plan to buy dedicated Bitcoin mining equipment in the near future.
Instead, the company plans to place new infrastructure at existing sites so power can be redirected between mining Bitcoin and powering AI tasks, depending on which activity is more profitable at the time.
Premium gym chains are entering a “golden era” that is ending or already in decline, as rising operating costs collide with shifting consumer preferences toward more flexible, community-based ways to exercise. Long-term memberships are shrinking, margins are pressured by higher rents and facility expenses, and competition from smaller, more personalized…
