In March 2026, US-listed company Cango Inc. released its first full-year financial report as a Bitcoin miner. The results revealed a striking contradiction: while total annual revenue soared to $688.1 million, the company reported a net loss of $452.8 million.
This dramatic contrast—rising revenue but deepening losses—quickly became a hot topic in the crypto market. Alongside its hefty loss report, Cango announced it had sold off most of its Bitcoin holdings and was making a decisive pivot to the artificial intelligence infrastructure sector. This move is more than a single company’s strategic choice; it reflects a broader industry-wide transformation as Bitcoin miners, facing post-halving cost pressures and a relentless computing power race, are undergoing a profound collective shift. In this article, we’ll use Cango’s financials as a lens to break down the structural causes behind its losses and examine the industry’s evolving narrative from "hashrate sales" to "compute services."
"Year One" of Transformation: Revenue Up, Profits Down
According to Cango’s unaudited financial results, 2025—the company’s first full fiscal year as a dedicated Bitcoin miner—saw total revenue reach $688.1 million. Of this, Bitcoin mining contributed $675.5 million, accounting for over 98% of total revenue. By revenue scale alone, Cango’s transformation got off to a rapid start. However, this impressive top line stands in stark contrast to a net loss of $452.8 million.
Shortly after releasing its financials, Cango made a key announcement: to reduce debt and strengthen its balance sheet, the company sold 4,451 Bitcoins in February 2026, raising approximately $305 million in cash. These funds would be used to repay long-term related-party debt and support its new strategy—pivoting to AI infrastructure services via the EcoHash platform.
Two Major Pivots: From Auto Finance to AI Compute
Cango’s transformation has been nothing short of radical, with its core business making two major shifts in just two years.
- Before November 2024: Cango’s core business was automotive finance. Seeking new growth, the company began divesting legacy assets in preparation for entering the crypto industry.
- November 2024: Cango officially entered Bitcoin mining, acquiring 32 EH/s of hashrate with operations spanning North America, the Middle East, South America, and East Africa—its first major leap.
- Full year 2025: Cango went all-in on mining, producing a total of 6,594.6 Bitcoins. During this period, the company terminated its ADR program and transitioned to direct listing on the NYSE to enhance corporate governance transparency.
- February 2026: Even before announcing its annual loss, Cango acted early by selling off most of its Bitcoin holdings, signaling a shift in capital management from "hold" to "liquidate."
- March 2026: With the annual report release, Cango officially declared its second leap to the market: evolving from a pure Bitcoin miner to an integrated energy and AI computing infrastructure platform.
Where Did the $452.8 Million Loss Come From?
To understand Cango’s losses, it’s not enough to look at revenue alone—you have to dig into its cost structure and asset impairments. Here’s a summary of key financial data from the report:
| Financial & Operational Metrics | Full Year 2025 | Q4 2025 | Key Drivers |
|---|---|---|---|
| Total Revenue | $688.1M | $179.5M | Driven primarily by mining operations |
| Net Loss | -$452.8M | -$285M | Costs and expenses far outstrip revenue |
| Bitcoins Mined | 6,594.6 BTC | 1,718.3 BTC | ~18 BTC mined per day |
| All-in Mining Cost | $97,272/BTC | $106,251/BTC | Includes depreciation and all operating expenses |
| Mining Equipment Impairment | $338.3M | $81.4M | Driven by price volatility and equipment efficiency reviews |
| Bitcoin-Backed Receivables FV Loss | $96.5M | $171.4M | Revaluation of related-party collateralized loans |
A review of these numbers reveals that the core drivers of loss are a combination of high costs and asset write-downs:
- Cost Inversion: The all-in mining cost for the year reached $97,272 per Bitcoin, and soared to $106,251 per Bitcoin in Q4. According to Gate market data, as of March 18, 2026, the Bitcoin price was about $74,492.2. This means that in Q4, Cango lost over $30,000 on every Bitcoin it mined—a classic "mine one, lose one" scenario that directly drove the losses.
- Massive Asset Impairments: Total annual operating costs and expenses reached about $1.1 billion. Mining equipment impairment alone accounted for $338.3 million, reflecting the shrinking value of mining rigs as fixed assets under price pressure and rapid tech evolution. At the same time, Bitcoin price volatility led to nearly $100 million in losses from fair value changes in "Bitcoin-backed receivables."
Market Perspectives: Strategic Breakout or Forced Retreat?
Cango’s massive loss and subsequent AI pivot have sparked a range of interpretations in the market.
- Official Company View: Strategic Costs and Investments
Cango’s management frames the loss as a result of "non-recurring transformation costs" and "market-driven fair value adjustments." CFO Michael Zhang emphasized that the company is lowering leverage by adjusting its Bitcoin reserve strategy and reallocating capital for new ventures like AI. CEO Paul Yu painted an optimistic future, leveraging Cango’s scale in compute and energy networks to deliver cost-effective AI inference services through EcoHash. In this narrative, losses are the price of transformation, and AI is the ultimate goal.
- Market & Analyst View: Unsustainable Model Triggers Emergency Pivot
External observers focus more on the mining sector’s underlying challenges. Quinn Thompson, CIO of Lekker Capital, noted that the deteriorating mining economics are an "underappreciated headwind" for the Bitcoin market. For high-cost miners like Cango, the old business model is no longer viable.
The prevailing market view sees the Bitcoin sell-off and AI pivot as a forced "self-rescue." AI hosting can deliver more stable and higher-margin revenue streams than Bitcoin mining. Cango’s stock has dropped over 84% in the past six months and now hovers around $0.68, signaling deep investor pessimism about its mining prospects. Selling Bitcoin for cash is seen as a survival move—a necessary price for staying afloat and finding a new narrative when profits dry up.
Assessing the Narrative: Is the AI Pivot a Smooth Path or a Risky Bet?
Cango’s AI transformation story must be examined in the broader industry context, where both its credibility and its challenges are evident.
- Industry Trend Support: Cango is not alone. Core Scientific plans to sell "nearly all" of its Bitcoin holdings in 2026 to fund AI expansion; TeraWulf has made high-performance computing hosting its main growth engine, signing over $12.8 billion in long-term customer contracts. This shows that pivoting energy infrastructure from "mining" to "AI compute services" is a real and ongoing industry trend. These companies already have the power resources, sites, and operational expertise that AI data centers require.
- Transformation Challenges: However, shifting from mining to AI services is no easy feat. Cango faces several key hurdles:
- Funding Gap: Building AI infrastructure is capital-intensive, requiring much larger upfront investment than mining. The $305 million raised from Bitcoin sales is just "seed money"—future capital access will be critical.
- Technical Barriers: Facilities designed for Bitcoin mining need major upgrades to meet AI workloads’ stringent demands for network latency, cooling, and stability. The CEO’s mention of "site retrofitting" and "product development" are still in early stages.
- Competitive Landscape: Cango must not only compete with other pivoting miners but also face industry giants like Amazon Web Services and Google Cloud. In the AI compute rental market, latecomers must offer highly competitive pricing or unique technical advantages.
Scenario Analysis: Possible Evolution Paths
Based on the analysis above, Cango and similar miners may face several possible future scenarios:
- Scenario 1: Successful Transformation
If Cango can successfully retrofit its facilities, leverage the EcoHash platform to carve out a cost or efficiency advantage in a niche (such as edge AI inference), and secure major clients, it could unlock a second growth curve. In this case, the current losses and asset sales would be seen as a successful strategic repositioning, and its stock price could recover.
- Scenario 2: Dual Pressure
If the AI business underperforms and fails to generate enough cash flow, while the original mining business loses competitiveness due to asset sales and wavering strategy, the company could find itself stuck—unable to lead in AI and falling behind in mining. This would put sustained pressure on its balance sheet.
- Scenario 3: Mining Supply-Side Shakeout (Macro Scenario Affecting All Miners)
On a broader scale, the transformation or exit of high-cost miners like Cango and Core Scientific represents a "supply-side shakeout" in the Bitcoin network. Falling hashrate would ease competitive pressure for remaining miners, laying a healthier foundation for the next cycle. In this scenario, Cango’s individual outcome would be part of a larger industry adjustment.
Conclusion
Cango’s $452.8 million loss is more than a financial record—it’s a mirror reflecting the harsh realities of a maturing Bitcoin mining industry: scale no longer guarantees profit; cost control and asset efficiency are now the keys to survival. Its pivot to the AI sector is both a response to losses and a bet on future industry trends. The road ahead is full of uncertainty, but it highlights how "compute power" as an infrastructure asset is evolving—from simply minting "digital gold" to serving a more diverse digital economy. Whatever the outcome, Cango’s attempt has already left its mark on the history of crypto mining.
