Meta’s Deal for Nuclear Power Is Likely Cheaper Than Microsoft’s, Jefferies Says

Jefferies estimates Meta's new 20-year power agreement with the largest US nuclear operator for its AI operations will likely cost around $80 per megawatt hour from the Clinton plant in Illinois. This price point is projected to be more favorable than a comparable deal struck by Microsoft last year, reflecting the increasing demand and strategic importance of securing stable energy sources for AI infrastructure.

Analysis

Meta Platforms Inc. has entered into a significant 20-year nuclear power contract with the largest US nuclear operator to fuel its expanding artificial intelligence operations, a move indicative of the surging energy demand within the AI sector. Jefferies LLC analyst Paul Zimbardo forecasts Meta will likely pay approximately $80 per megawatt hour for energy from the Clinton plant in Illinois. This pricing is notably projected to be more advantageous than a comparable agreement secured by competitor Microsoft Corp. last year, potentially offering Meta a cost efficiency in a critical operational area. The strongly positive sentiment surrounding this deal for Meta underscores the strategic importance of securing stable, long-term, and competitively priced energy sources to support the intensive power requirements of AI infrastructure, positioning Meta favorably in managing a key input cost for future growth.

Key Decisions for Investors

• Investors should consider Meta's ability to secure this long-term power agreement at a potentially lower cost than Microsoft as a positive factor for its AI operational expenditure and competitive cost structure.
• For those invested in Microsoft, this development may highlight a comparative cost disadvantage in energy procurement for AI, warranting attention to Microsoft's future energy sourcing strategies and their impact on margins.
• The trend highlights the increasing materiality of energy costs for AI-driven companies; therefore, monitoring such long-term energy agreements and their terms will be crucial for assessing the scalability and profitability of AI investments across the sector.