Anthropic Needs Gigawatts of Power It Doesn't Have. This Dividend-Paying Industrial Sells It.

NextEra Energy Resources reported 18% adjusted EPS growth in its latest quarter and expanded its generation and storage backlog by 3.6GW to more than 35GW, positioning it to supply AI data-center demand through renewables, gas, storage and nuclear. The company secured up to $1.9B in DOE loans to restart its 615MW Iowa nuclear plant by early 2029. Management targets roughly 10% dividend growth this year, 6% annual dividend growth through 2028, and at least 8% annual EPS growth through 2035; the stock's recent decline has lifted its dividend yield to about 3.2%.

Analysis

NEE's AI-power optionality is less about a single data-center contract than its ability to bundle firm generation, storage, renewables and transmission into long-duration PPAs. That reduces the intermittency discount facing pure renewable developers, but the equity re-rating requires conversion of backlog into contracted projects with acceptable returns; a large pipeline alone is not a near-term earnings catalyst. The reopened Iowa unit is strategically useful for signaling firm-power credibility, yet its 2029 start date means it should not be capitalized as a material 2026-28 AI earnings contributor.

The more immediate valuation driver is financing: NEE's development model remains unusually sensitive to long-end Treasury yields, tax-credit monetization and project-level construction costs. A lower-rate cycle could expand the utility/renewables multiple and improve project economics over the next 3-12 months, whereas higher-for-longer rates or delayed interconnection/transmission approvals would pressure both growth and dividend-coverage perceptions. DOE financing lowers project funding risk but does not eliminate outage, refurbishment-cost, fuel-supply, and schedule risk at the nuclear asset.

Consensus may be overpaying for speculative nuclear and fuel-cell AI beneficiaries while underappreciating incumbent utilities' execution constraints. NEE is the higher-quality diversified expression, but its regulated utility capital program can compete with Energy Resources for balance-sheet capacity; watch whether incremental AI-related commitments raise the consolidated funding need faster than operating cash flow. The cleanest structural beneficiaries from grid congestion may also be transmission equipment suppliers such as ETN, PWR and HUBB, which monetize capacity additions regardless of which generation technology wins.

Key Decisions for Investors

• Accumulate NEE on rate-driven weakness over a 6-18 month horizon; size as a quality utility/AI-infrastructure exposure rather than a near-term nuclear trade. Thesis is invalidated by a material reduction in long-term EPS/dividend growth targets, sustained 10-year Treasury yields above the level that forces higher project financing costs, or evidence that backlog conversion is slowing.
• Use a relative-value expression: long NEE / short a basket of pre-revenue advanced-nuclear or fuel-cell AI proxies where available, with a 3-12 month horizon. NEE has contracted-asset and regulated-cash-flow support, while speculative peers are more exposed to permitting, commercialization and dilution; close if NEE's capital-spend or leverage outlook deteriorates materially.
• For a higher-beta, technology-agnostic AI-power buildout trade, favor ETN or PWR over a 6-18 month horizon rather than adding only generation exposure. Grid and electrical-equipment spending benefits from new load additions even if data centers source power from gas rather than renewables or nuclear; risk is a hyperscaler capex reset or transmission-permitting delay.
• Do not assign value to the Iowa restart beyond a probability-weighted long-dated option until refurbishment budget, offtake economics, and commissioning milestones are independently disclosed. Treat any sharp NEE rally solely on nuclear headlines as an opportunity to wait for contract and financing verification.