TeraWulf: The AI Infrastructure Transformation Is Taking Shape

TeraWulf has pivoted from bitcoin mining to AI data-center leasing, securing more than ~$27B in long-term contracted revenue across 839MW of critical IT load. Key commitments include a 20-year, 401MW Anthropic lease and Fluidstack/Google-backed agreements; annual rent is projected at ~$1.76B at full delivery. Earnings are expected to remain negative until rents ramp in 2028, while the stock trades at approximately 11–12x all-in EV/stabilized EBITDA.

Analysis

The key market risk is not tenant demand but conversion: WULF’s value depends on delivering powered capacity on schedule and funding construction before the contracted rents begin. The quoted stabilized multiple can look attractive while obscuring years of execution, financing and dilution risk; long leases reduce demand uncertainty, not build-out risk. A delay in power or fit-out could push cash flows further out and compress the multiple before any tenant default is relevant.

The contract headlines also need entity-level diligence. Verify which party is legally liable for rent, what guarantees or security support the obligations, and whether the stated revenue is fixed rent or conditional on delivery and acceptance. In particular, do not treat “Google-backed” as equivalent to an Alphabet guarantee: the article does not establish that Alphabet is the contracting obligor. No direct fundamental read-through to GOOG is warranted absent evidence of material economics or liability.

Near term, milestones on financing, construction and power delivery are more important than the headline contracted-revenue figure. Over 1–3 months, disclosed project budgets, funding terms and schedule updates could reprice WULF; over 6–18 months, successful delivery would support its transition from a bitcoin-mining valuation to AI infrastructure. The contrarian risk is that investors capitalize the 2028 rent ramp too early: a signed lease is not stabilized EBITDA. Conversely, if project funding is secured on acceptable terms and delivery milestones are met, the market may be underweighting the scarcity value of powered capacity.

Key Decisions for Investors

• Keep WULF on a catalyst watchlist rather than buying solely on the stabilized-EBITDA multiple. Reassess after verifying project-level capex, committed financing, dilution risk, power-delivery dates and lease commencement conditions.
• If those checks are favorable, consider a small, staged WULF long around independently confirmed construction and energization milestones; define the thesis as successful conversion to rent-paying capacity, not simply continued AI demand.
• Do not use GOOG as a direct hedge or short against WULF on the “Google-backed” wording. First establish whether Alphabet has a material contractual obligation; absent that, the exposure is not demonstrated.
• Falsifiers: material schedule slippage, funding on substantially dilutive or costly terms, reduced contracted capacity, or lease payments contingent on milestones that WULF fails to meet. A shift from projected rent toward actual delivered, rent-paying capacity would strengthen the thesis.