AI can move at software speed. Electricity cannot. Investors are paying handsomely for the difference.
At $437.28, Eaton Corporation presents me with an unusual problem.
I think the business is better than it has ever been. I am considerably less convinced the shares are.
Eaton has transformed itself from a diversified industrial manufacturer into a critical infrastructure supplier behind AI data centres, electrification, utility investment and factory reshoring. Revenue has climbed from $19.6 billion in 2021 to $30.0 billion trailing twelve months, while free cash flow has risen from $1.59 billion to $3.93 billion.
Wall Street has noticed. At roughly 29x forward earnings, investors are no longer valuing Eaton as an ordinary industrial company.
They are paying what I call the gridlock premium: an elevated valuation for owning scarce electrical capacity precisely when the world cannot build power infrastructure quickly enough.
Does that scarcity permanently belong in Eaton’s multiple? Some of it, but not all — which is why I anchor my valuation closer to 25x than 29x.
AI moves instantly. The electrons are stuck in traffic
A $15 Billion Reservation Book
Electrical Americas backlog reached approximately $15.2 billion at June 2026, up 33% year on year, while rolling twelve-month orders increased 41% organically. Across Eaton’s combined Electrical businesses, rolling twelve-month book-to-bill stood at 1.2.
Normally, backlog is simply revenue waiting to happen. Here, Eaton’s manufacturing slots, engineering expertise and qualified equipment increasingly resemble reserved capacity.
A hyperscaler cannot casually replace critical switchgear, power distribution and protection systems midway through a multi-billion-dollar campus. Certification, engineering integration and reliability matter.
Scarcity therefore brings pricing power as well as visibility — and Electrical Americas’ 27.5% second-quarter operating margin shows that strong demand is translating into attractive economics.
Bulls argue this structural scarcity deserves a higher multiple; bears argue 29x assumes scarcity and pricing power persist long after competitors expand capacity. The investment decision comes down to what the current price already assumes.
My first warning light would be combined Electrical book-to-bill falling below 1.0 without an obvious temporary explanation. That would suggest shipments are outrunning replacement orders and scarcity may be normalising.
The Grid Can Be Customer and Roadblock
Insufficient electrical infrastructure creates demand for Eaton while simultaneously constraining the projects buying its equipment.
Connecting huge new loads requires substations, switchgear, transformers, protection systems, energy-storage integration and increasingly microgrid architecture. Yet a data-centre campus awaiting utility capacity can postpone later phases, shift orders or redesign around available megawatts.
A $15.2 billion backlog is impressive; it is not $15.2 billion sitting in the bank.
The grid is therefore both Eaton’s opportunity and its speed limit. Scarcity can increase electrical content and pricing, while interconnection delays postpone the projects creating that demand. For investors paying a substantial multiple, delayed demand can make quarterly expectations considerably less cooperative.
What $437 Actually Requires
This is where I become cautious.
Eaton expects 2026 adjusted EPS of approximately $13.40–$13.60. Suppose I demand a 10% annualised return over five years and, for simplicity, apply a 25x terminal earnings multiple to 2031 EPS.
The maths requires approximately $28 of EPS in 2031. Ignoring the dividend, Eaton needs to compound earnings at roughly 16% annually from the 2026 guidance midpoint.
Even if Eaton retains approximately 29x earnings in 2031, EPS would still need to compound around 12–13% annually to produce that return.
That is what “priced for perfection” means here: considerable future growth is already capitalised into today’s price.
Next-twelve-month consensus EPS of approximately $15.08 puts the shares near 29x. At 25x, the same earnings produce about $377 a share — roughly 14% below today’s price.
Financial Deep Dive: Growth Now Has Debt Attached
Eaton’s operating transformation is genuine. Since 2021, revenue has risen more than 50%, operating income has more than doubled from $2.5 billion to roughly $5.5 billion trailing twelve months, and free-cash-flow margin has expanded from 8.1% to 13.1%.
The multiple expanded, but the financial machine improved first
Then Eaton spent $9.5 billion acquiring Boyd Thermal.
Boyd adds liquid-cooling technology and extends Eaton’s data-centre proposition towards the chip itself. As AI racks become more power-dense, electricity and thermal management are converging into one infrastructure problem.
Strategically, I like it. Financially, $Eaton Corp PLC(ETN)$ paid approximately 22.5x Boyd’s estimated 2026 adjusted EBITDA.
Total debt increased from approximately $10.7 billion at year-end 2025 to $21.3 billion, leaving net debt around $20.6 billion — just above 3x trailing EBITDA. Trailing ROIC has meanwhile fallen from 15.8% in 2025 to approximately 10.6%.
That is my second warning light. If ROIC does not recover as Boyd’s earnings enter Eaton’s numbers, I would question whether management bought growth faster than it created value.
Nearly $4 billion of trailing free cash flow provides meaningful deleveraging capacity, but dividends and capital investment also compete for that cash. At 22.5x EBITDA, Boyd needs to perform.
The Hyperscaler Question Nobody Should Ignore
Eaton’s greatest opportunity also contains its third warning light.
Data-centre demand ultimately depends on a relatively small universe of hyperscalers, colocation operators and developers. Eaton does not disclose a clean data-centre percentage for Electrical Americas sales, but in Q1 2026 data-centre orders increased approximately 240% year on year and revenue about 50%.
That is spectacular growth — and evidence of how powerfully this end market is influencing Eaton’s incremental trajectory.
$Microsoft(MSFT)$, $Amazon.com(AMZN)$, Alphabet and Meta need not stop building for expectations to wobble. A collective moderation in capex growth or prolonged project deferrals could be enough. I would therefore watch hyperscaler capital-spending guidance almost as closely as Eaton’s own orders.
The customer may not appear prominently on Eaton’s invoice. Economically, it can still be sitting at the other end of the cable.
Eaton Engineered Its Own Re-Rating
I resist dismissing Eaton’s valuation as AI exuberance because management deliberately created a better business before investors awarded it a better multiple.
The Hydraulics divestment, earlier portfolio reshaping and planned Mobility separation have progressively concentrated Eaton around higher-growth, higher-margin Electrical and Aerospace businesses.
The market is not simply paying more for yesterday’s Eaton. Some of the re-rating is deserved; I simply don’t think the full 29x should be treated as permanent.
Portfolio surgery has worked. The patient now looks considerably fitter — and considerably more expensive.
Eaton Versus the Power Crowd
Schneider Electric, ABB and Siemens ensure Eaton is no monopoly. Its advantage lies instead in North American scale, installed relationships, engineering qualification and breadth across critical electrical infrastructure.
Beyond the equipment manufacturers, $Quanta(PWR)$ provides another useful comparison from the grid-buildout side of the same infrastructure boom. Its transmission, substation and large-load infrastructure work sits upstream of much of the electrical demand Eaton ultimately serves.
Vertiv provides the sharper comparison.
Vertiv trades around 31x forward earnings, only modestly above Eaton’s roughly 29x. Yet Vertiv expects 30–32% organic sales growth and 58–61% adjusted EPS growth in 2026, while carrying much greater direct exposure to digital infrastructure.
On a growth-adjusted basis, $Vertiv Holdings LLC(VRT)$ arguably looks cheaper: investors pay only around two additional turns of earnings for substantially faster growth.
Its greater AI concentration cuts both ways, however. Eaton offers data-centre upside alongside utilities, commercial buildings, aerospace and industrial electrification.
I therefore anchor Eaton around 25x, roughly a 20% discount to Vertiv. The precise discount is a judgement, not a formula: Eaton’s slower growth warrants it; its broader diversification prevents me demanding more.
Eaton trades at almost Vertiv money while supplying considerably more seat belts. I want those seat belts; I just want a discount for the slower car.
At 25x, that is not bargain-bin pricing. Nor should it be.
Valuation sets my price; volume shows where investors actually committed
I Want Eaton, Just Not at Any Voltage
My position is straightforward.
At roughly $377, I would start building a position provided the operating thesis remains intact. At $437, I would hold existing shares but not add meaningfully until the valuation moves towards that threshold. Around $400, I would still wait unless higher forward earnings had brought the multiple materially closer to 25x.
Combined Electrical book-to-bill below 1.0, Boyd-era ROIC failing to recover, or broad hyperscaler capex retrenchment would make me reconsider the thesis. Continued double-digit electrical growth, backlog expansion, improving ROIC and rapid deleveraging could justify a higher entry valuation.
That matters because $377 is not sacred. It is today’s expression of 25x forward earnings. If earnings expectations rise materially while those indicators remain healthy, I would raise the entry price with them.
A great future still has a price today
Eaton has become one of the most strategically important industrial companies in the AI economy. But today’s investor is being asked to pay today for a sizeable portion of tomorrow.
The power shortage is real. The moat is real.
At $437, so is the price.
Comments