Cango's $285M Loss: A Flow Analysis of Mining Economics and the AI Pivot

The financial flow in Q4 was starkly negative. Revenue of $179.5 million was completely overwhelmed by operating costs and expenses, which surged to $456.0 million. This massive gap directly led to a net loss from continuing operations of $285 million for the quarter.

The core problem was the cost of production. While $172.4 million of that revenue came from mining Bitcoin, the total cost to mine each coin was $106,251. This all-in cost exceeded the average Bitcoin price during the period, meaning the company lost money on every Bitcoin it produced. The result was a quarterly operating loss of $276.6 million, even before accounting for other significant charges.

The bottom line was a severe cash burn. The $285 million net loss, driven by the fundamental mismatch between mining revenue and costs, highlights the extreme pressure on the business model during the quarter. This flow of money out of the company, despite scaling production, underscores the vulnerability of its operations to Bitcoin's price volatility and its own high cost structure.

The Balance Sheet Catalyst: Selling Bitcoin to Survive

The company's dire cash position forced a direct financial catalyst. In February 2026, Cango executed a sale of 4,451 BTC for approximately $305 million. This move was explicitly to reduce debt and free capital to build an AI inference business, a pivot that management says strengthens the balance sheet.

The scale of this sale is striking. It consumed a significant portion of the company's full-year 2025 production, which was 6,594.6 Bitcoin. In other words, Cango sold nearly 68% of its annual output in a single transaction just months after its first full year of mining. This is a classic liquidity event, converting a key asset into cash to fund a strategic shift.

From a market flow perspective, this adds a large, discrete block of Bitcoin to the circulating supply. While the sale was planned and disclosed, the sudden injection of over 4,000 coins can contribute to selling pressure, potentially pressuring the price. It also signals that the mining cash flow, which was already negative, is insufficient to fund the new AI ambitions.

The AI Pivot and Market Context: A New Compute Narrative

Cango's pivot to AI inference is a direct financial response to unprofitable mining. The company's net loss of $285 million in Q4 2025 and its sale of 4,451 BTC in February to fund the shift underscore the failure of its core mining model. Management's stated goal of building an AI business is a strategic retreat from a sector where costs have become crippling.

The broader market context shows this pivot is a high-stakes bet on a new narrative. Miners that diversified into AI and HPC saw triple-digit stock returns in 2025, dramatically outperforming pure-play miners. Companies like IREN and Cipher posted returns of 285% and 218% respectively, fueled by major deals with tech giants. This performance gap creates immense pressure for miners like Cango to replicate that story, as investors are clearly rewarding the shift to higher-margin compute.

The high-cost environment makes this pivot necessary but also risky. The sector-wide pressure is illustrated by Marathon Digital's average mining cost of $67,704 per Bitcoin in Q3 2025. Cango's own cost was even higher at $106,251 per coin. In this environment, the AI narrative isn't just about growth-it's about survival, offering a potential path to a more stable and profitable revenue stream away from volatile Bitcoin prices.