Canaan (CAN) Q1 2026 earnings review
Landmark Order Hangover: Q2 Guidance Cuts Deeper Than the 'Trough' They Promised
Q1 revenue of $62.7M landed in line with Canaan's own guidance, down 68% sequentially as the massive North American order that drove Q4 2025 finished shipping. But the bigger signal is Q2 2026 guidance of $35-45M—well below Q1 and roughly half of pre-shock baseline quarters. Management told investors in February that Q1 would mark the bottom; the new guide says otherwise. Gross profit swung to a $22.9M loss as inventory write-downs hit $24.5M (vs $13.9M in Q4 and $2.5M a year ago). Cash fell from $80.8M to $43.5M, partly a timing issue—$42M was collected in April—but accounts receivable nearly tripled to $51.6M. The bright spot: a smart all-share deal for 49% of the ABC mining sites in West Texas at sub-$0.03/kWh power, growing the crypto treasury to a record 1,808 BTC / 3,952 ETH, and disciplined cost cuts (OpEx -18% YoY).
🐂 Bull Case
Acquired 49% of three operating West Texas mining sites from Cipher Mining via share exchange—no cash outlay. The sites carry ~4.4 EH/s of operational hashrate and 120 MW of installed power capacity at sub-$0.03/kWh, among the lowest cost positions in the industry. Hashrate at the JV has already grown to 4.82 EH/s by end of April.
Operating expenses fell 18% YoY and 11% QoQ to $31.4M. G&A down to $15.0M (from $16.9M Q4), R&D down 19% YoY. Mining business continues to generate positive cash contribution at $0.04/kWh all-in power cost even through the BTC price plunge.
🐻 Bear Case
In February, management positioned Q1 as the cyclical low. Three months later, they're guiding Q2 revenue 36% below Q1 at the midpoint—and roughly half of any pre-2025-Q3 baseline quarter. This is not stabilization; it's continued deterioration.
Without the $24.5M inventory write-down, gross margin would be 2.5% (effectively breakeven). Inventory write-downs have escalated three quarters in a row: $2.5M → $13.9M → $24.5M. Adjusted EBITDA loss widened to $76.3M from $40.5M in Q4.
⚖️ Verdict: 🔴
Bearish. The execution on a known-bad quarter was clean, but Q2 guidance broke the trough narrative, and underlying margins are barely breakeven before write-downs. The ABC deal and energy strategy are credible long-dated assets, not near-term offsets to cash burn of ~$35M/quarter on operations and another ~$40M/quarter in crypto fair-value swings.
Key Themes
Q2 Guide Contradicts the 'Bottom' Narrative
Three months ago, management told investors Q1 2026 would be the low point and gave a $60-70M range for the quarter. Q1 came in at $62.7M—near the low end of that range. But the Q2 guide of $35-45M (midpoint $40M) implies a further 36% sequential decline. Computing power sold collapsed from a record 14.6 EH/s in Q4 to just 4.1 EH/s in Q1, and ASP fell back to $10.50/TH from $11.30/TH in Q4 and $11.80/TH peak in Q3 2025. Both volume and pricing are decelerating. CEO Zhang's own framework: machine demand recovery typically appears when hash price reaches $40-45/PH/day; it's currently 'around $30-some dollars' and recently dropped further. The hardware business is not at a bottom—it's still falling.
Escalating Inventory Write-Downs Signal Underlying Demand Weakness
Inventory and prepayment write-downs have grown three quarters running: $2.5M (Q1 25) → $13.9M (Q4 25) → $24.5M (Q1 26). The Q1 write-down alone was nearly 40% of total Q1 revenue. This is the company marking down inventory to a market value that keeps falling. CFO Cheng framed it as 'continuous pricing pressure and aligned our inventory cost structure with the market environment'—candid translation: we keep being too optimistic about what this inventory is worth. Excluding the write-down, Q1 gross margin would be roughly 2.5%—essentially breakeven.
Cash Position Tight, AR Tripled
Cash fell from $80.8M to $43.5M during the quarter. The company emphasized that $42M was collected in April—true—but the corresponding accounts receivable balance ballooned 167% from $19.3M to $51.6M, meaning collection on Q1 product sales was deferred to a single concentrated payment. This installment policy for major customers is a real customer-financing concession. Adjusting for the April collection, effective liquidity is ~$85M, but Adjusted EBITDA loss was $76.3M for the quarter (including $40.9M in non-cash crypto fair-value losses; 'operating' Adj EBITDA was ~-$35M). At current operating burn, runway is finite without further financing.
ABC Project: Smart Capital-Light Pivot to Energy Infrastructure
The biggest strategic asset added this quarter. Canaan acquired Cipher Mining's 49% interest in three operating West Texas mining sites (Alborz, Bear, Chief Mountain) for $25M in stock—54M ADSs at fair value, zero cash outlay. The sites bring ~4.4 EH/s of operational hashrate at end of Q1 (4.82 EH/s by end of April after fleet upgrades with HQ) and 120 MW of installed capacity at below $0.03/kWh power cost. The deal also delivered 6,840 Avalon A15 Pro mining units now on Canaan's books as PP&E. Cipher became a significant shareholder, aligning the two companies. This is the kind of deal you want to see in a downturn: real operating assets with low power cost, acquired with paper trading at a discount.
Mining Operations Resilient Despite Bitcoin Price Drop
Total installed hashrate reached 11 EH/s, up 11% QoQ and 66% YoY. The company produced 257 BTC in Q1 generating $19.1M of mining revenue, with an all-in power cost of ~$0.04/kWh. BTC production fell from 300 in Q4 to 257 here despite higher hashrate because of weather-related curtailments in North America (the company voluntarily powered down during winter storms) and the steep drop in Bitcoin price. Despite the price decline, mining operations continued to generate positive operating cash contribution. The crypto treasury grew to record levels: 1,807.6 BTC and 3,951.5 ETH, with combined market value of ~$140M at recent prices.
Cost Discipline Showing Through
Total operating expenses fell to $31.4M in Q1, down 18% YoY and 11% QoQ. G&A came down to $15.0M (vs $16.9M both Q4 and Q1 2025), driven by a $2.1M sequential reduction in staff cost. R&D fell 19% YoY to $15.4M, primarily on lower staff and share-based comp. Sales & marketing down 59% YoY. Management also called out deeper AI tool adoption across R&D, supply chain, and finance functions. This is genuine cost reduction during a downturn—not the lip-service kind.
Macro: Bitcoin Price Crash Reset the Cycle
Bitcoin entered the year near $95k, bottomed near $66k in early March, ending Q1 at ~$67k (down 23% in the quarter). Hash price collapsed to around $30/PH/day. The Middle East geopolitical situation, tightening global liquidity, and elevated energy costs all weighed on miner confidence. CEO Zhang's own demand framework: machine market only meaningfully recovers when hash price reaches $40-45, and 'goes crazy' at $55. Hash price has been climbing slowly toward $40 but recently dropped back. CEO noted last year's BTC highs were partly USD-weakness-driven rather than a 'typical breakout cycle'—a notably sober framing from the CEO of a mining hardware company.
A16 Series Ready, Demand Isn't
The A16 XP launched in Q4 2025 with 300 TH/s output and 12.8 J/TH efficiency. In Q1, sample units shipped to select customers for testing, with reportedly strong feedback on stability, energy efficiency, and noise control. Mass production is ready. The problem isn't product—it's that customers globally remain in 'wait-and-see' mode due to compressed economics. Management is intentionally not building inventory aggressively, taking a 'lean inventory' posture instead. The A16's $30k breakeven shutdown price (vs $37k for A15 Pro) gives it cycle-resilience when demand returns, but the timing of that return is the question.
Nordic Hash-to-Heat: New Use Case Validating Energy-Compute Strategy
Won a competitive bid to provide hash-to-heat infrastructure to a Nordic district heating network using Avalon A1566HA hydro-cooled units. Total planned deployment is 8 MW, with 2 MW already operating and supplying hot water to residents. Customer placed a follow-on order in March 2026 for an additional 6 MW after the initial deployment. Small in dollar terms but strategically important: it validates Canaan's water-cooling technology and demonstrates a non-mining use case for the hardware. CEO described this as the kind of fragmented but standardizable opportunity the company can replicate.
Other KPIs
Decelerating sharply. Worsened from -$40.5M in Q4 25 and -$38.1M in Q1 25. The Adjusted EBITDA reconciliation includes $40.9M of non-cash crypto fair-value losses (BTC fell from ~$87k to ~$67k during the quarter). Excluding those, 'operating' Adjusted EBITDA was approximately -$35M—still a meaningful step worse than the modest profitability of Q2 and Q3 2025, reflecting the collapse in product volume and the inventory write-down.
Accumulating steadily. BTC holdings grew from 1,749.9 at year-end 2025 to 1,807.6 at end of Q1 (+57.7 BTC net of any sales/usage). Combined market value at recent prices (~$77k BTC) is approximately $140M—roughly equal to the company's current cash + post-quarter collections. 905 BTC are pledged as collateral for secured term loans and 100 BTC are in fixed-term products. The treasury has become a meaningful balance sheet component, but also creates fair-value volatility that flows through Adjusted EBITDA each quarter.
Avalon Home (consumer) generated $2.7M in Q1, 6.3% of product revenue. North American customers contributed over 80% of product sales, up from 75% in Q4—but this is a denominator effect from the collapse of the broader product base, not a sign of strength. The company is actively expanding home product distribution via Best Buy Canada and Amazon, but at $2.7M/quarter this segment is not yet material to consolidated results.
Guidance
Decelerating. Midpoint of $40M implies a 36% sequential decline from Q1's $62.7M and a 60% YoY decline from Q2 2025's $100.2M. This is the most important number in the release: management had positioned Q1 as the trough in February's call, but the Q2 guide undercuts that thesis. CEO cited continued miner caution, BTC price volatility, geopolitical uncertainty, and energy price pressure. Notably absent: any directional guidance on H2 2026 recovery, FY revenue, or breakeven targets.
Key Questions
Is the Cycle Bottom Still in Sight?
In February you positioned Q1 as the trough of this cycle. Q2 guidance points 36% lower. What specifically broke since then—customer cancellations, ASP, channel inventory, BTC price—and what would have to change for you to call a bottom with any confidence now?
Cash Runway and Financing Options
With $43.5M cash at quarter-end, $42M collected in April, and operating Adjusted EBITDA losses around $35M per quarter, runway looks roughly two to three quarters at current burn before requiring action. What's the priority order: additional equity issuance via ATM (paused since 2025), using crypto treasury as collateral, debt, or further cost cuts? Why has the buyback continued at this cash level?
Customer Concentration and AR Risk
Accounts receivable nearly tripled this quarter to $51.6M against Q1 product revenue of $42.9M, implying nearly all product sales were extended on installment terms. How many customers does that exposure concentrate among, and what protections (collateral, escrow, deposits) sit behind it?
ABC Project Economics for Canaan
The 49% equity stake in ABC Projects represents real operating cash flow at sub-$0.03/kWh power. What was the JV's net income contribution in Q1 ($0.2M equity-method gain was disclosed but that's a partial quarter), and what's the run-rate contribution we should expect on a full-quarter basis at current BTC prices?
AI/HPC Strategy: Concrete Milestones?
You've now talked about AI/HPC infrastructure for multiple consecutive quarters but have explicitly declined to share site counts, capacity by stage, or signed agreements. What specific commercial or contractual milestone would unlock more disclosure—a signed PPA, a customer LOI, a permit? What's the realistic timeline?