
CleanSpark reported fiscal 2025 revenue of $766 million (+102% YoY), net income of $364.5 million (EPS $1.25), adjusted EBITDA >$823 million, produced ~8,000 BTC and holds >13,000 BTC (~$1.2 billion), and completed an upsized $1.15 billion zero‑coupon convertible note to fund a $460 million buyback and shore up working capital (~$1 billion). By contrast, TeraWulf posted Q3 revenue of $50.6 million but a GAAP net loss of $455 million driven by non‑cash warrant/derivative revaluations, carries nearly $1.5 billion of debt against $712.8 million cash (negative net cash ~$374 million), faces preferred conversion dilution (141.9483 common shares per preferred) and negative free cash flow (~$35 million in Q3), while pursuing an ambitious AI pivot with targeted 200–250 MW HPC by year‑end 2026 and potential Fluidstack contracts. The piece concludes CleanSpark is materially better positioned on profitability, cash generation, mining efficiency and pragmatic AI deployments, whereas TeraWulf carries elevated leverage, dilution and execution risk.
Market structure: The sector is bifurcating—well‑capitalized, operationally efficient miners that can pivot to AI colocation (CleanSpark/CLSK) are winners; highly levered, cash‑burning miners (TeraWulf/WULF) are losers. CleanSpark’s 50+ EH/s, $1.2bn BTC treasury and $1.15bn zero‑coupon convert give it optionality to monetize 285MW Texas capacity; WULF’s ~$1.5bn debt vs $712.8m cash (net cash ≈ -$374m), $0.035/kWh guidance and mandatory preferred conversion (141.9483:1 in Dec‑2025) compress its pricing power and raise dilution risk. Expect pricing power to accrue to operators with long‑term low‑cost contracts and modular immersion capability (benefiting Submer, NVDA indirectly), while power‑sensitive marginal miners will be price takers. Risk assessment: Key tail risks—sudden BTC price shock (<$30k) that forces asset sales; regulatory intervention on crypto/energy use; failed WULF execution on Abernathy requiring >$5bn capex; and convertible dilution shocks for CLSK if equity conversion economics change. Time horizons: days—elevated equity and option IV around Dec‑2025 preferred conversion and any BTC swings; weeks–months—funding milestones, AI customer announcements (priority window: next 90 days); quarters–years—AI revenue ramp to 2027 and Bitcoin cycle recovery. Hidden dependency: AI revenue depends on hyperscaler procurement cadence and grid interconnection timelines, not just acreage or announced MW. Trade implications: Primary trade: establish a 2–3% long position in CLSK within 2–6 weeks (target +30–50% in 12 months if BTC>=$50k and 100–200MW of AI bookings confirmed), stop‑loss 20% below entry. Pair trade: long CLSK / short WULF (equal dollar exposure) to isolate sector‑AI vs leverage risk; for WULF prefer buying 3–6 month put spreads (limited capital, horizon through Dec‑2025 conversion) or shorting on rallies above 20% from current levels. Options: sell covered calls on CLSK after partial size if premium >10% 30‑day; buy WULF ATM 3M puts if preferred conversion settlement date approaches. Contrarian angles: Consensus underestimates that WULF’s Fluidstack/Google linkage could be structured as low‑capex revenue share—if so, upside may be underpriced; conversely, market may be under‑discounting latent dilution from WULF preferred conversion and derivative revaluations. Historical parallel: 2017 miners that chased hosting often failed on execution and grid limits—watch interconnection queue positions and firm power start dates as true leading indicators. Unintended consequence: rapid AI conversion raises localized power prices and regulatory scrutiny, which could compress margins across the cohort even for efficient operators.
AI-powered research, real-time alerts, and portfolio analytics for institutional investors.
Request a DemoOverall Sentiment
moderately positive
Sentiment Score
0.35
Ticker Sentiment