Canaan (CAN) Q1 2026 Earnings Transcript

Canaan reported Q1 2026 revenue of $63 million, in line with guidance, but gross loss of $23 million reflected a $25 million non-cash inventory write-down and a $41 million fair value loss on digital assets. Product revenue was $43 million with over 80% from North America, while mining revenue was $19 million and cash-positive despite weak Bitcoin pricing. Management guided Q2 revenue to $35 million-$45 million and highlighted a strategic pivot toward energy infrastructure and AI/HPC, supported by the share-based ABC project acquisition.

Analysis

The important read-through is not the revenue miss versus the prior cycle; it’s that CAN is actively transforming from a hardware vendor into a call option on scarce power. That matters because in this market the scarce asset is no longer ASIC design alone, but grid access with permissive regulation and low-cost load optionality. A share-based acquisition of operating power and infrastructure is economically attractive only if management believes its equity is cheap relative to embedded power value; that is a tacit signal that the stock is still trading below replacement value for the strategic footprint it is assembling.

The second-order effect is that CAN’s mining operation is becoming a balance-sheet and operating-data generator for the next business model, not just a profit center. If they can arbitrage load between mining and future HPC workloads, the company may end up with a utilization advantage that pure miners and pure HPC developers both lack. The catch is timing: the asset base is being built now, but monetization of the AI/HPC angle is likely measured in quarters-to-years, while the market is still trading the name on hash price and hardware ASPs over the next 1-2 quarters.

Near term, the main catalyst is not Bitcoin price alone but the slope of hash price recovery and whether product demand snaps back enough to absorb inventory without more write-downs. If hash price stays subcritical, the equity remains a financing-sensitive, high-beta levered play on a weak end market. The contrarian point is that consensus is probably underestimating how much of the downside has already been absorbed by the cleanup actions; if BTC stabilizes and hash economics firm into the high-$40s per TH/day equivalent, demand could inflect faster than bears expect because the company has kept inventory light and preserved operating capacity.

Biggest risk is that the market keeps rewarding AI narrative names while punishing all crypto-linked hardware, leaving CAN stranded between categories. That creates a valuation gap that can persist unless management shows a tangible monetization event in power/HPC or a sustained rebound in machine orders over the next 1-2 quarters.

Key Decisions for Investors

• Long CAN as a tactical, high-beta recovery trade, but only on weakness and with a 4-8 week horizon; the setup improves if BTC holds above the mid-$70k zone and hash price trends toward the mid-$40s threshold that historically reopens miner ordering.
• Pair trade: long CAN / short a basket of weaker crypto-miner proxies with poorer balance sheets and less North American power access over the next 1-2 quarters; CAN has a clearer path to survive the downcycle via infrastructure ownership rather than pure hardware exposure.
• Buy 1-2 month CAN call spreads instead of outright stock if you want the BTC rebound convexity without full downside exposure; the stock remains sensitive to another inventory or asset-marking shock if crypto rolls over again.
• Fade the AI-HPC re-rating near term: do not pay up for a full AI multiple until management converts the pipeline into disclosed, financeable assets; treat the narrative as a long-dated option, not a near-term earnings driver.
• If CAN rallies on a BTC bounce without a corresponding order recovery, take profits into strength; the risk/reward worsens quickly because the market can reclassify the name back to a cyclical miner when the narrative outruns cash flow.