Vistra: AI Power Provider Meets Load Growth & Acquisition Upside

VST expects FY2027 adjusted EBITDA of $7.6B-$8.3B, with potential upside from Cogentrix and Meta PPAs, while maintaining leverage of roughly 2.3x. Regulatory moratoriums could delay projects, but VST's existing ERCOT and PJM asset base, PPA pipeline, colocation strategy, and Helix Digital Infrastructure partnership position it to serve multi-year AI data-center power demand.

Analysis

VST’s differentiated value is not simply generation ownership; it is the ability to monetize existing interconnection, transmission access, and dispatchable capacity near high-load demand. That scarcity can support contracted-price premiums and reduce development risk relative to greenfield power projects, but it also makes VST unusually exposed to permitting or interconnection-rule changes that strand incremental load growth. The key earnings sensitivity over the next 12-24 months is the conversion of announced data-center demand into signed, creditworthy, long-duration contracts—not preliminary partnership announcements.

The market may be underestimating the margin trade-off in hyperscaler PPAs. Contracts with META-like counterparties lower volume and credit risk, potentially supporting a higher earnings multiple, but fixed-price structures can cap upside if ERCOT/PJM capacity prices, gas volatility, or nodal congestion intensify. Conversely, a demand slowdown would affect merchant power pricing and contracted-load additions simultaneously, creating more downside beta than a conventional utility. Cogentrix integration is a secondary catalyst only if it adds dispatchable capacity or commercial optimization capability without increasing leverage or maintenance-capex requirements.

Near term, regulatory headlines can create tradable volatility but are unlikely to alter cash flows unless they directly affect load interconnection, colocation approvals, or market-capacity design. Over 6-18 months, VST’s rerating depends on demonstrating that incremental AI load raises EBITDA per MW rather than merely increasing capital intensity. Falsify the constructive thesis if net leverage trends above roughly 3x, contracted-load milestones slip by two quarters, or ERCOT/PJM forward power and capacity curves weaken materially despite AI-demand announcements.

Key Decisions for Investors

• Maintain a tactical long VST only on regulatory-driven weakness, sized as a 6-12 month position; require evidence of executed PPAs, identified MW/term, and no material deterioration in leverage. Target a 15-20% upside from contract de-risking versus an 8-10% stop if forward power/capacity curves roll over.
• Express the cleaner relative-value view as long VST / short XLU over 6-12 months, rather than an outright utility-sector long: VST has greater scarcity exposure to data-center load, while regulated utilities face heavier rate-base funding and political-return constraints. Exit if VST’s contract conversion cadence fails to exceed sector load-growth expectations.
• Do not add META exposure on this information alone. Monitor whether data-center power commitments raise META’s long-term infrastructure spend faster than monetization of AI products; a material capex-guide increase without corresponding advertising or AI revenue signals would be a separate 1-3 month margin-risk catalyst.
• Set alerts around ERCOT and PJM interconnection, colocation, and capacity-market rulings. A rule that restricts large-load connections or shifts upgrade costs to generators would impair VST’s scarcity premium and warrants reducing exposure before revised EBITDA guidance.