Aterio Flags Delay Risk for Oracle's 1.3 GW Wisconsin Data Center

Aterio's review of regulatory filings through September 24, 2026 indicates that Vantage Data Centers' Project Lighthouse campus in Port Washington, Wisconsin, being built for Oracle, faces a material risk of missing Oracle's second-half 2027 delivery target. The analysis suggests meaningful power load at the site is unlikely before mid-2028, potentially delaying Oracle's planned infrastructure capacity expansion.

Analysis

The relevant transmission channel is not a one-quarter construction variance but Oracle's ability to convert contracted AI/cloud demand into recognized OCI revenue. If incremental capacity is unavailable when customers expect it, Oracle can retain bookings while deferring consumption revenue, creating a widening RPO-versus-revenue gap, weaker cloud growth optics, and potential multiple compression at the next two earnings prints. The cost side is also unfavorable: pre-committed power, networking, and accelerator-related obligations can begin ahead of billable load, pressuring OCI gross margin and free-cash-flow conversion.

This is a relative issue for ORCL rather than a broad data-center demand signal. Microsoft, Amazon, Google, and CoreWeave can redirect workloads across larger geographic footprints; Oracle's more concentrated capacity buildout makes a single delayed campus disproportionately important if it was allocated to a major customer cohort. Suppliers with diversified exposure to grid equipment and power management, including ETN, VRT, GEV, and PWR, are unlikely to see material demand destruction—delays more likely shift delivery timing than eliminate spend.

The contrarian case is that Oracle can bridge capacity through leased third-party facilities, reallocate workloads, or prioritize higher-value AI customers, making the revenue impact smaller than implied by a physical-site timeline. The key falsifier is management disclosure: sustained OCI growth and margin guidance without an increase in capex, lease commitments, or deferred implementation commentary would indicate that capacity substitution is working. Treat the third-party analysis as an alert, not confirmation, until permitting, utility interconnection, and Oracle capacity commentary can be independently verified.

Key Decisions for Investors

• Do not add directional ORCL exposure solely on this report; establish an earnings watch item for OCI growth, RPO growth, capex, and lease commitments over the next 1-3 months. A material OCI guide-down or widened RPO-to-revenue conversion gap would validate a tactical short.
• If ORCL has rallied into the next earnings release without revised capacity guidance, consider a defined-risk bearish structure: buy 3-6 month ORCL put spreads, sized for a 5-10% post-earnings downside rather than an outright short. Exit if OCI growth guidance is maintained or raised and management identifies substitute capacity.
• For a relative expression after verification, long MSFT versus short ORCL over 3-6 months captures Oracle-specific execution risk while reducing broad AI infrastructure beta. Close the spread if Oracle demonstrates workload reallocation or if Microsoft reports its own material power/capacity constraints.
• Maintain constructive exposure to diversified electrical-infrastructure suppliers ETN, VRT, GEV, and PWR rather than treating a project slip as a demand warning. Reassess only if multiple hyperscaler customers simultaneously defer equipment orders or utility-interconnection delays broaden beyond a single site.
• Validate whether VNTG is an investable public security before taking any position; absent confirmed liquidity, listing status, contractual economics, and project-specific revenue exposure, it should remain a research watch rather than a trade.