Nuclear Stock Face-Off: Is Constellation Energy or Vistra the Better Buy Right Now?

Constellation Energy’s nuclear revenue visibility is supported by 20-year power agreements with Microsoft (835MW Crane restart) and Meta (1,121MW Clinton), with additional 920MW PPAs signed in 2Q (through 6/30/2026). Constellation expects base EPS to compound 20%+ annually through 2029, though base EPS is only ~60%-70% of adjusted operating earnings, while Vistra has 20-year Meta (2,609MW) and AWS (up to 1,200MW) contracts and targets 2027 adjusted EBITDA of $7.4B-$7.8B. Despite Constellation trading at a higher multiple (22.4x vs Vistra 14.4x forward one-year earnings), the article frames Vistra as offering the stronger risk-reward due to the valuation gap and Vistra’s ~30% share count reduction since 11/2021.

Analysis

The equity market is pricing nuclear supply as a long-duration contracted cash-flow stream, but the real differentiator is not installed megawatts; it is how much of the future cash flow is already locked, how much remains optional, and how much buyback capacity can amplify EPS. That makes the lower-multiple name more interesting only if the market is willing to underwrite execution risk around asset uptime, contract conversion, and acquisition integration. CEG’s premium is defensible, but once a stock trades like a scarcity asset, any delay or hiccup in restart/permitting can trigger fast multiple compression over the next 1-3 months.

The clearest second-order winners are the hyperscalers, but the benefit is strategic de-risking rather than near-term P&L accretion; secure power matters most if AI buildouts are power-constrained in a handful of regions. The more durable spillover is into the uranium/fuel-cycle complex and transmission/interconnection names, while merchant gas peakers and other dispatchable but uncontracted generators face longer-term share loss as load gets pre-committed. Over 6-18 months, the key question is whether new supply and slower AI capex eventually normalize power scarcity and compress these “AI utility” multiples.

Contrarian take: the crowd may be underestimating how much of the apparent growth is already financed by financial engineering, especially for the stock with heavier reliance on buybacks and add-on deal optionality. If power demand softens or regulators slow nuclear execution, the cheaper name can still lag because leverage cuts both ways. The thesis fails if contracted load keeps expanding faster than expected, or if guidance revisions show that the market is still underappreciating 2027-2029 earnings power.

Key Decisions for Investors

• Long VST / short CEG as a 3-6 month relative-value trade if the market keeps paying for contracted visibility at any price; target 10-15% relative outperformance for VST, but stop if CEG announces another material hyperscaler contract or VST execution risk widens.
• If you want second-order AI power exposure, buy MSFT or AMZN on dips rather than chasing CEG/VST; the upside is option value from supply security, not direct EPS lift, so size as a small satellite position with a 6-12 month horizon.
• Watch, don’t buy yet, the uranium/fuel-cycle basket (e.g., UUUU) for a 6-18 month follow-through trade; only become aggressive if more utility contracting tightens spot pricing and enrichment lead times, which would validate the scarcity thesis.
• Set a tactical alert on the CEG/VST multiple spread: if the cheaper name remains sub-15x while 2027/2029 earnings visibility improves, add; if the spread narrows without fresh contracts, take profits because the rerating is already in the tape.