Galaxy Announces Second Quarter 2026 Financial Results

Galaxy Digital reported Q2 2026 net loss of $85M (diluted and adjusted EPS: -$0.09), primarily tied to depreciation in digital asset prices, while adjusted gross profit rose to $43M and adjusted EBITDA was -$77M. On the business side, Galaxy advanced its AI data center buildout—delivering 133MW of critical IT load to CoreWeave under a 15-year lease and securing $3.5B of senior secured notes due 2031 to fund Helios Phase II. Equity ended at $2.7B and cash & stablecoins were $2.46B (down 6% QoQ), signaling pressure from crypto market moves but continued infrastructure execution.

Analysis

The market should separate the two stories here: the trading/capital-markets business is still a cyclical crypto beta with noisy marks, while the data-center piece is becoming a longer-duration cash-flow annuity. The equity is likely to re-rate only if investors believe the incremental project cash flow is large enough to offset higher leverage and execution risk; otherwise the stock remains a financing story masquerading as an infrastructure story.

Second-order winners are the custody/staking intermediaries and adjacent institutional rails: BNY and STT can capture wallet-share from tokenization and staking workflows without taking construction or power-approval risk. By contrast, CRWV benefits from dedicated capacity, but the concentration also creates a single-counterparty dependency that can compress Galaxy’s negotiating leverage over time if the tenant knows the asset is bespoke.

Catalyst path is split by horizon. Over days, the print may be treated as positive because investors like visible contracted revenue. Over 1-3 months, the real test is whether interest expense, interconnect timing, and lease ramp convert the headline into repeatable EBITDA; any slip would reclassify the story as levered project finance. The contrarian miss is that the market may be underpricing how much of the upside is already monetized by the balance sheet expansion and how little is left for common equity once debt service and buildout capex are deducted.

Key Decisions for Investors

• Short GLXY on post-earnings strength or via 1-3 month put spreads; thesis is that the market is overpaying for long-dated AI optionality while underweighting leverage, counterparty concentration, and ERCOT execution risk. Falsifier: no delay in Phase II, lower-than-expected interest burden, and clear pre-leasing momentum.
• Pair trade: long CRWV / short GLXY for 1-3 months. CRWV owns the compute-demand scarcity with cleaner operating leverage, while GLXY carries the funding and buildout overhang. Risk/reward improves if the market keeps rewarding pure AI exposure over infrastructure developers.
• Buy STT or BNY on weakness as the cleaner institutional digital-asset infrastructure beneficiary. These names can monetize custody/staking and tokenization flow without capital-intensive real-asset execution; use as a lower-volatility proxy versus GLXY.
• If staying long GLXY, wait for the conference call and only add after management quantifies Q3 lease run-rate and financing costs. The trade works only if the market sees a credible path from project cash flow to equity compounding.