HIVE Digital (HIVE) Q4 2026 earnings review

Record Hashrate Growth Overshadowed by Severe Margin Squeeze

HIVE successfully executed a massive operational expansion in FY26, quadrupling its Bitcoin mining hashrate to 25.1 EH/s and driving full-year revenue up 158%. However, Q4 laid bare the harsh realities of the crypto cycle: as Bitcoin prices cooled and network difficulty spiked, top-line revenue decelerated sequentially, and Adjusted EBITDA flipped negative (-$9.0M). To escape this relentless mining treadmill, management is aggressively accelerating its pivot toward High-Performance Computing (HPC) and AI, announcing plans for a colossal 320 MW AI Gigafactory in Toronto to chase $660M in recurring revenue.

๐Ÿ‚ Bull Case

HPC Pivot is Materializing

The deployment of the liquid-cooled NVIDIA B200 cluster at Bell Canada immediately lifts HPC ARR to $35M. Targeting >$200M ARR by year-end 2026 provides a visible path to de-risking the business from crypto volatility.

Execution of 25 EH/s Scale-up

Management proved it can build. The 300 MW hydro-powered expansion in Paraguay was delivered on time, pushing HIVE into the top tier of global, green-energy miners.

๐Ÿป Bear Case

Hashprice Reality Check

Despite having 4x more computing power than a year ago, Q4 mining revenue fell 24% sequentially. If Bitcoin prices stagnate while global hashrate rises, the mining unit will continue to bleed cash.

Gigafactory Funding Gap

The planned 320 MW Toronto Gigafactory requires CAD $3.5 billion in capital. For a company that just raised $115M in convertible notes and posted negative Q4 EBITDA, financing this ambition without massive dilution is highly questionable.

โš–๏ธ Verdict: โšช

Neutral. The operational execution in building the Paraguay site and scaling the NVIDIA clusters is excellent. However, the core mining business economics are sharply deteriorating, and the capital requirements for their ambitious AI transition are daunting.

Key Themes

DRIVER ๐ŸŸข

Paraguay Expansion Delivered

HIVE successfully scaled its mining operations from 6.5 EH/s to 25.1 EH/s in just one year, driven primarily by completing three phases of a 300 MW hydro-powered buildout in Paraguay. This allowed the company to mine 104% more Bitcoin YoY (2,885 BTC), vastly outpacing the 42% increase in global network difficulty over the same period.

DRIVER NEW ๐ŸŸข

Accelerating HPC Cloud Segment

HIVE is rapidly migrating from a crypto-miner to an AI infrastructure provider. Post-quarter (May 2026), HIVE activated its first 504 NVIDIA B200 liquid-cooled GPU cluster with Bell Canada, signing contracts at $2.90 per GPU-hour (32% above forecast). This single deployment boosted contracted Annual Recurring Revenue (ARR) from $20M to $35M.

CONCERN NEW ๐Ÿ”ด

Q4 Margin Squeeze Exposes Mining Fragility

Management champions the 4x hashrate growth, but Q4 data contradicts the victory lap. Reversing from a highly profitable Q3, Q4 Adjusted EBITDA turned negative (-$9.0M), and Gross Operating Margin compressed to 24% (down from 49% in Q2). A 27% sequential drop in average Bitcoin price combined with a 27% YoY surge in network difficulty easily wiped out the benefits of scale.

CONCERN NEW ๐Ÿ”ด

Staggering Depreciation Burden

HIVE reported a massive $148.4M GAAP net loss for the year, largely driven by $170.4M in depreciation. While management rightfully notes this is a non-cash charge, it highlights the severe capital intensity and short functional lifespan of ASIC mining rigs, which require constant replacement to remain competitive.

THEME โšช

Capital Structure Evolution via Exchangeable Notes

In April 2026, HIVE shifted its funding strategy, closing $115M in 0% Exchangeable Senior Notes due 2031 (exchangeable at ~$2.57/share). While this avoids current cash interest, it introduces significant potential equity dilution, which the company partially hedged with $19.8M in capped-call transactions.

CONCERN NEW ๐Ÿ”ด

The CAD $3.5 Billion Elephant in the Room

HIVE announced the acquisition of land for a 320 MW AI Gigafactory in Toronto, estimating a capital requirement of CAD $3.5 billion to reach full capacity (100,000 GPUs). For a company generating $73M in annual Adjusted EBITDA, bridging this monumental funding gap presents severe execution and dilution risks.

Other KPIs

FY26 Gross Operating Margin $107.9 million

A 329% increase YoY, expanding the margin profile to 36% for the year (up from 22% in FY25). This was driven heavily by the operating leverage of the 300 MW Paraguay facility and the ~40% margin profile of the growing HPC business. However, investors must monitor the Q4 run-rate (24%), which shows severe recent degradation.

Q4 General & Administrative Expense $9.4 million

G&A crept up sequentially from $8.4M in Q3, pushing full-year G&A to $31.4M (nearly double the $16.6M in FY25). Management attributes this to scaling headcount for the Paraguay operations and expanding the BUZZ HPC platform.

Guidance

HPC Annual Recurring Revenue (ARR) - CY 2026 >$200 million

Accelerating aggressively. Management expects to double its GPU cloud fleet from ~5,500 to ~11,000 NVIDIA GPUs by the end of calendar 2026. If achieved, this will radically alter the company's revenue mix away from Bitcoin.

GTA Gigafactory Long-Term ARR Target $360 million

HIVE estimates that its planned 320 MW Toronto facility, assuming 200 MW of IT load and $150 per kW/month colocation pricing, could generate $360M in annualized recurring revenue. This project is targeted to come online in the second half of 2027.

Key Questions

Financing the Gigafactory

The Toronto GTA Gigafactory carries an estimated CAD $3.5 billion price tag. Beyond the recent $115M note issuance, what specific project financing, hyperscaler partnerships, or debt instruments will HIVE utilize to fund this without destroying shareholder equity?

Bitcoin Break-even and CapEx Deferment

With Adjusted EBITDA turning negative in Q4 due to hashprice compression, at what Bitcoin price does HIVE begin curtailing operations or pausing further ASIC refresh cycles in favor of preserving capital for the HPC business?

HPC Capacity Bottlenecks

Scaling to 11,000 GPUs by the end of the calendar year requires not just chip procurement, but significant power, cooling, and network infrastructure. What are the primary supply chain bottlenecks for achieving this target, and how much of the necessary CapEx is already committed?