Bitfarms Surges 7.14% on AI Pivot Despite 417th Market Activity Rank

On September 24, 2025, Bitfarms (BITF) closed with a 7.14% gain, despite a 35.89% drop in trading volume to $0.24 billion, ranking it 417th in market activity. The stock has surged 73.83% year-to-date but remains below its 2021/2022 highs. CEO Ben Gagnon highlighted the company’s strategic pivot from bitcoin mining to high-performance computing (HPC) and AI infrastructure at the H.C. Wainwright conference, emphasizing long-term cash flow from HPC/AI contracts while maintaining exposure to BTC’s upside. Current operations are expected to sustain stable free cash flow through 2026, leveraging low-cost energy and operational efficiency.

Bitfarms reported Q2 2025 revenue of $78 million, up 87% year-over-year, though gross mining margins fell to 45% from 51% in Q2 2024. Administrative expenses rose to $18 million, driven by the integration of Stronghold Digital. The firm sold 1,052 BTC at an average of $95,500 in Q2, generating $100 million in proceeds. Gagnon outlined plans to expand HPC/AI infrastructure at its Panther Creek campus, positioning the company near AI hubs like Amazon and CoreWeave. With over 1 GW in Pennsylvania energy projects and strategic assets in Washington and Quebec, Bitfarms aims to capitalize on North America’s AI infrastructure boom.

Comparative analysis shows Bitfarms has a lower price-to-sales ratio (5.21) than peers Hut 8 (6.98) and Marathon Digital (8.67), suggesting a value-oriented play. The firm’s vertically integrated energy portfolio and disciplined BTC sales strategy contrast with Marathon’s scale-driven approach and Hut 8’s high-growth focus. However, margin pressures from rising electricity costs and execution risks in the AI pivot remain challenges. Political support for data center development in key regions and proximity to major tech players could bolster market share in the AI infrastructure sector.

The proposed backtest strategy involves selecting the top 500 most liquid stocks daily, purchasing them at the next session’s open, and exiting at the following close. Implementation requires defining a tradable proxy, such as an ETF (e.g., SPY, VTI) or a predefined basket, due to current tool limitations in cross-sectional ranking. The performance period spans from January 1, 2022, to the present, with results dependent on the selected proxy and execution methodology.