Value Destruction in Bitcoin Microchip Manufacturers: A Market Anomaly?

In early 2021, Canaan Creative (CAN), a major manufacturer of Bitcoin mining equipment, traded above $36. Today, the stock trades around $0.40, a decline of roughly 99%. Ebang International (EBON) tells a similar story. Over the same period, Bitcoin has more than doubled. 

But the share price alone can be misleading for companies that issue large amounts of new equity. Over roughly five years, Canaan increased its ADS count – American Depositary Shares, which allow foreign companies to trade on US exchanges – by approximately 4.4x, to 690.6mn from about 166mn. 

Even after adjusting for dilution, the market value destruction remains severe: Canaan's equity value fell to approximately $280mn from roughly $5.5bn.

Meanwhile, Canaan's fourth-quarter 2025 revenue reached $196.3mn, its highest quarterly level in about three years. At first glance, that looked like an operational recovery. The most recent data tell a different story. In Q1 2026, revenue collapsed to $62.7mn, down 68% sequentially, while net loss widened to $88.7mn. For Q2 2026, management guided between $35 and $45mn, with a midpoint 59% below analysts' consensus of $98mn. The apparent recovery stopped before it could prove itself.

(Source: Tradingview)

The value chain fracture

The Bitcoin mining value chain has three layers: Bitcoin microchip manufacturers, such as Canaan, Ebang, and the private players Bitmain and MicroBT, miners that buy and operate those machines, such as MARA Holdings (MARA), Riot Platforms (RIOT) and CleanSpark (CLSK), and Bitcoin itself. In theory, all three layers should rise and fall together. In practice, the chain has fractured.

(Source: Tradingview)

Listed miners are not all abandoning Bitcoin mining, but many are trying to monetize their access to power, sites and capital through AI (AI), high-performance computing (HPC) and data-centre contracts.

For Bitcoin microchip manufacturers, the AI exit route is not entirely closed, but it is far less direct. An AI data centre primarily uses Nvidia or AMD GPUs, not SHA-256 ASICs (application-specific integrated circuits designed exclusively for Bitcoin mining). Canaan tried to develop an AI edge-computing semiconductor business, but formally discontinued it in June 2025.

Is the ASIC industry dead?

If listed miners are shifting towards AI and data centers, who is still buying mining machines? 

The answer is visible in the hashrate — the total computing power securing the Bitcoin network. Bitcoin's network hashrate recently reached around 1 ZH/s, showing that the ASIC market is not dead, but its customer base is changing. The buyers are no longer primarily the large listed US miners. 

They increasingly include private miners, energy-rich jurisdictions, sovereign actors, and a new retail/home-mining category.

(Source: Coinmetrics)

Canaan now reports three main revenue streams: industrial mining equipment, home mining equipment, and self-mining. Industrial B2B equipment sales remain dominant but extremely volatile. Self-mining has become a second revenue base. Home mining is still small but strategically important because it opens a retail channel in an industry that was historically almost entirely B2B.

Avalon Home, Canaan's consumer product line that includes Bitcoin-mining heaters and compact home miners, represents a genuine sector breakthrough. The company has also expanded into hash-to-heat and grid balancing projects in the Nordics, Japan and Canada, where waste heat from mining machines is reused for district heating or power-grid support.

(Source: Canaan Quarterly Results)

The ASIC industry is not dead, as Bitcoin hashrate remains near historic highs. However, the customer base is shifting away from large listed US miners towards private, sovereign, energy-integrated, and retail buyers.

Canaan is not the moribund company implied by its share price. Instead, it is undergoing a profound transformation, supported by three reported revenue segments and several adjacent initiatives across home mining, self-mining, heat reuse, modular hash boards and energy-integrated compute.

Risks nevertheless remain elevated, with a Q1 2026 net loss of $88.7mn, ongoing dilution, structurally dominant competition from Bitmain, and recurring exposure to inventory write-downs.

The investment case is ultimately binary. If Avalon Home, hash-to-heat, self-mining and energy infrastructure reach critical scale before dilution overwhelms shareholders, the current valuation could appear anomalous in hindsight. 

Conversely, if losses persist at the current pace, further dilution is likely to erode most of the remaining upside.