Cango reported a $452.8 million net loss in 2025 despite generating $688.1 million in revenue, underscoring cost pressures in large-scale bitcoin mining. The results highlight how capital intensity and balance sheet volatility continue to weigh on new entrants.
Cango Inc.
The NYSE-listed firm generated $675.5 million, or over 98% of revenue, from mining after entering the sector in November 2024 with 32 exahashes per second (EH/s) of computing power. Total operating costs reached $1.1 billion, with fourth-quarter expenses alone at $456 million, including impairment and fair-value losses tied to bitcoin collateral.
Are Bitcoin Miners Shifting Capital Toward AI Infrastructure?
Cango produced 6,594.6 bitcoin (BTC) in 2025, averaging 18.07 BTC per day, with output rising slightly in the fourth quarter. Yet rising production did not offset cost escalation, reflecting a broader pattern seen across the mining sector where efficiency gains are often matched by higher capital and energy expenditures.
The company attributed losses partly to non-recurring transformation costs and valuation adjustments. CFO Michael Zhang said the firm is adjusting treasury policy and securing equity funding to manage volatility, while reducing leverage during its transition phase.
Cango has already begun reallocating capital. It sold approximately $305 million in bitcoin post-year-end, cutting reserves by 60% to fund expansion into artificial intelligence infrastructure.
“We are advancing our pivot to become an AI infrastructure provider,” said CEO Paul Yu, citing plans to deploy compute capacity for AI inference through its EcoHash platform.
This shift mirrors a wider trend among public miners. Core Scientific has indicated it may sell most of its bitcoin holdings in 2026 to finance AI and high-performance computing, while MARA Holdings recently expanded its policy to allow reserve sales despite holding 53,822 BTC worth $4.7 billion at year-end.
The convergence of mining and AI compute is becoming more pronounced as firms seek higher-margin revenue streams. Market participants are now watching whether AI deployments can deliver more stable cash flows than bitcoin mining, particularly as energy costs and hardware impairments continue to pressure profitability.
