Wall Street increasingly has two versions of NextEra Energy.
The bull sees an AI-power compounder: America needs enormous quantities of electricity, grid connections take years, and NextEra owns one of the deepest development pipelines in the industry. The bear sees something less glamorous: a heavily indebted utility spending almost $30 billion a year on capital projects while issuing debt and shares to keep the machine moving.
I think both arguments deserve attention. The investment question is which version ultimately earns the higher valuation.
At $76.83, $NextEra(NEE)$ is neither a nuclear momentum trade nor a sleepy income utility. It is increasingly a wager on whether the scarce asset in American power is not generation technology itself, but deliverable megawatts: land, interconnection, permits, transmission, generation and somebody capable of assembling them before the data-centre customer loses patience.
The AI race has reached a gate measured in megawatts
The Queue Is the Moat
America can manufacture more turbines, panels and batteries. It cannot manufacture five years overnight.
Interconnection delays have turned existing grid access into an increasingly valuable asset. NextEra Energy Resources’ approximately 35.1 GW renewables and storage backlog therefore matters beyond its headline size. FPL has also reported roughly 21 GW of large-load interest, including around 12 GW in advanced discussions.
There is evidence that hyperscalers are engaging rather than merely browsing. NextEra’s partnership with Google includes the development of multiple gigawatt-scale data-centre campuses, combining new generation and load development.
But the numbers still need an asterisk large enough to power a small server rack. Interest is not contracted revenue. Advanced discussions are not operating data centres. AI infrastructure spending could slow, projects could be cancelled, and demand remains concentrated among a relatively small number of hyperscalers with considerable bargaining power. The opportunity is enormous, but the certainty is not.
NextEra Is Selling Time
The old NextEra bull case was straightforward: FPL supplied regulated growth, while NEER supplied renewable growth. Add a rising dividend and stir gently.
The new model is more interesting. Storage adds dispatchability, gas provides firm capacity and nuclear has returned through the planned restart of Iowa’s Duane Arnold plant. NextEra is therefore not betting the company on renewables defeating nuclear; it is building an all-of-the-above development machine.
AI data centres do not particularly care whether Wall Street prefers uranium, gas or batteries. They care whether several hundred megawatts arrive on schedule and remain available.
NextEra is increasingly selling speed-to-power.
Financial Deep Dive: Growth Has an Electricity Bill
TTM revenue is $28.7 billion, up 10.8%, while operating income is $8.44 billion. Reported net income is $9.30 billion and diluted EPS $4.45.
That unusual relationship deserves attention. Net income exceeding operating income suggests headline GAAP earnings are benefiting from significant below-operating-line items. Consequently, the 17.3× trailing P/E makes NEE look cheaper than its underlying operating economics necessarily justify. Management’s adjusted EPS guidance is the cleaner valuation anchor.
NextEra expects 2026 adjusted EPS of $3.92–$4.02, targeting the upper end, while continuing to target 8%+ annual adjusted EPS growth through 2032. At $76.83, the shares trade at roughly 18.8× consensus forward earnings, based on forward EPS of approximately $4.09, and yield 3.25%. That is hardly an extravagant AI multiple.
The balance sheet is where things become considerably less relaxing.
TTM operating cash flow is $13.8 billion against $29.8 billion of capital expenditure, implying conventional free cash flow of approximately negative $16.0 billion. Total debt has risen to $110.2 billion from $55.0 billion in 2021, with net debt around $107.3 billion. Debt-to-EBITDA stands at approximately 7.55×.
The opportunity grew quickly. The balance sheet grew with it
Negative free cash flow is not inherently alarming for a utility building regulated and contracted assets; those investments should produce future cash flows. But financing them still costs real money. NextEra issued a net $17.9 billion of debt over the TTM period and roughly $2.0 billion of common stock.
Higher long-term rates therefore attack the thesis twice: they increase financing costs while making NEE’s 3.25% dividend less competitive with lower-risk income alternatives.
Even the dividend offers a clue. Management expects roughly 10% annual dividend growth through 2026, slowing to around 6% from year-end 2026 through 2028. Capital allocation is becoming more demanding.
FPL: Shock Absorber With a Regulator Attached
FPL generated $18.72 billion of TTM gross revenue versus $9.53 billion at NEER. Its regulated earnings give NextEra stability while the development business spends heavily.
Yet regulated does not mean risk-free. FPL’s current settlement uses a 10.95% authorised ROE, below the 11.90% originally requested. As electricity demand and infrastructure spending rise, regulators must balance grid investment against household affordability.
That tension could intensify if residential customers perceive themselves as subsidising infrastructure built for enormous corporate data centres. FPL’s large-load tariff structure is designed to protect existing customers, but future rate cases will test whether the economics work as neatly in practice as they do in PowerPoint.
Regulators, inconveniently, do not value PowerPoint animations.
Competitive Analysis: Four Ways to Sell Scarcity
$Constellation Energy Corp(CEG)$ offers the cleaner nuclear proposition: established 24/7 carbon-free generation attractive to hyperscalers, with less dependence on constructing an enormous new renewable pipeline. Its roughly 20.8× forward multiple reflects some of that scarcity value.
$Vistra Energy Corp.(VST)$, at roughly 13.6× forward earnings, offers greater exposure to merchant generation and electricity prices. Its substantial hedging provides near-term cash visibility, but its earnings remain more exposed to power-market economics than NextEra’s regulated base.
$GE Vernova Inc.(GEV)$ sells the hardware enabling everyone else’s expansion. Turbines and grid equipment give it exposure to the power shortage without requiring it to finance entire generation portfolios.
NextEra sits between them. Its regulated operations provide earnings insulation that merchant generators lack, while NEER supplies development growth and exposure across renewables, storage, gas and nuclear. The price is considerably greater capital intensity and leverage.
That makes NEE less of a generation-technology bet than a coordination bet. Its advantage is the ability to solve several power bottlenecks simultaneously.
Price shows the move; volume reveals where investors actually committed
Dominion: Bigger, Safer — And Harder
The proposed all-stock $Dominion Resources(D)$ combination changes the investment case materially.
If completed, the combined company would derive more than 80% of its business mix from regulated operations, while management expects immediate adjusted-EPS accretion and 9%+ annual adjusted EPS growth through 2032. Those are attractive targets, not guarantees.
The transaction still requires multiple regulatory approvals before its expected second-half 2027 closing. NextEra must then integrate a vast multi-state utility organisation while delivering promised efficiencies and maintaining regulatory relationships. Because consideration is equity-based, shareholders must also consider dilution and how the enlarged share count interacts with future earnings growth.
Scale can lower procurement and financing costs. It can also turn a fast ship into an aircraft carrier.
What Would Make Me Wrong?
Three developments would undermine my thesis. First, if large-load ‘interest’ fails to become contracted, financeable demand, the megawatt moat is worth less than advertised. Second, if Dominion integration consumes management attention or expected accretion fails to emerge, greater scale becomes greater complexity.
Third, persistently high interest rates, weaker renewable economics or adverse tax and energy-policy changes could raise the cost of NextEra’s extraordinary investment programme just as its balance sheet is being asked to work hardest. I would also watch FPL regulatory outcomes closely: a development opportunity is considerably less valuable if regulators prevent shareholders from earning an adequate return on the capital required to serve it.
Growth sits on one side. Capital keeps the beam level
Buy the Bottleneck, Respect the Bill
At roughly 18.8× forward earnings, I think NEE offers an interesting compromise between utility defensiveness and AI-infrastructure growth.
My valuation discipline matters, though. Above roughly 22× forward adjusted earnings — about $90 using consensus forward EPS of approximately $4.09 — I would stop being constructive. That multiple would already exceed CEG’s roughly 20.8×, so NEE would need stronger contracted data-centre demand, higher earnings expectations or both to justify the premium through its regulated earnings base and broader development optionality.
The real moat is not renewable energy. It is not nuclear energy either. It is the ability to look across the table at a hyperscaler asking for several hundred megawatts and say: 'Yes, we can actually power that.'
In today’s electricity market, that may be the most valuable sentence in the room.
So how should NEE be valued: as a utility, an infrastructure developer, or something between the two? Does Dominion make NextEra safer, or simply larger and slower? And would you rather own NEE’s diversified megawatt machine, nuclear-heavy CEG or merchant-heavy VST?
Comments