Galaxy Provides Update on ERCOT Batch Zero Large Load Classifications

Galaxy Digital received conditional ERCOT Batch Zero classifications for five Texas data-center projects totaling approximately 4.2 GW of gross power capacity, including 1.63 GW at Helios classified as Base Load and remaining expansion projects classified as Studied Load. The company’s Texas development pipeline now exceeds 5.7 GW across four sites, positioning it for AI/HPC hosting expansion; Helios I and II remain scheduled for energization in 2028 with leasing unaffected. However, the classifications remain subject to ERCOT and Texas PUC audits, capacity allocation, and potential revisions, leaving interconnection timing and approvals uncertain.

Analysis

The market should value this primarily as a conditional real-option expansion rather than incremental contracted EBITDA. GLXY retains both upside from AI/HPC lease pricing and downside from retail-power procurement, meaning its eventual margin depends on the spread between customer power pass-through terms, ERCOT congestion/basis costs, and capacity-utilization—not merely on megawatts awarded. Until tenant commitments, construction capital requirements, and delivered-power economics are disclosed, the expansion is more likely to support narrative/multiple optionality than near-term estimates.

The key second-order risk is that Texas policy scrutiny raises the cost and duration of large-load development across the state. That would favor operators with already energized, transmission-secured campuses and penalize developers whose valuations embed speculative pipeline MW; likely relative beneficiaries include existing AI-hosting capacity at APLD, IREN and CIFR, although each has separate execution and crypto-beta risks. For GLXY, the more material downside is a regulatory delay that converts a differentiated power footprint into years of carrying costs and diluted development capital.

Near term, conditional treatment may produce a modest positive reaction because it reduces binary exclusion risk, but the next 1-3 month catalyst is audit/study clarity rather than leasing. Over 6-18 months, the stock can re-rate only if management demonstrates that capacity converts into creditworthy, long-duration AI contracts without excessive balance-sheet leverage. Contrarian view: investors may be underpricing the scarcity value of controllable retail-power agreements, but are likely overpricing gross MW until interconnection sequencing and tenant economics are independently verified.

Key Decisions for Investors

• Maintain a watch-list long bias in GLXY, but do not add materially on this release alone; initiate only after ERCOT confirms final status and GLXY discloses at least one binding AI/HPC lease or project-level funding framework. A favorable final determination without commercial commitments is insufficient for a fundamental underwriting.
• For existing GLXY exposure, use any event-driven strength to reduce crypto-linked beta and retain a smaller infrastructure optionality position. Falsification trigger: a withdrawal/downgrade of the higher-value classifications, a material delay to the existing energization timetable, or development funding that requires significant equity issuance.
• Consider a 3-6 month relative-value basket: long GLXY versus short a diversified proxy of pre-revenue Texas AI-data-center developers only after final regulatory clarity. The thesis is that GLXY's operating-power-control model can command better economics than pure land-and-power pipeline valuations; avoid the pair if GLXY cannot show customer deposits, contracted MW, or power-cost pass-through protections.
• Monitor ERCOT congestion pricing, PUCT large-load rules, and announced transmission upgrades as the highest-frequency thesis indicators. Sustained adverse basis/congestion moves or rules requiring new grid-cost contributions would impair projected hosting margins and should override the long thesis even if capacity classifications are upheld.