Okay, so here’s a fun one. The AI trade has basically moved through three phases so far. First it was chips. Then it was memory and networking. And now the market is quietly starting to ask a question that feels almost too obvious: what happens when the actual bottleneck isn’t silicon at all, but electricity?

That question leads us straight to GE Vernova (NYSE: GEV), the energy spinoff that split off from General Electric about two years ago. If you’re not familiar with it, no worries, most people aren’t thinking of this as an “AI stock” yet, and that’s kind of the whole point of this piece.

What GE Vernova Actually Does

GE Vernova runs three segments: Power (gas turbines, nuclear, hydro), Electrification (grid transformers, switchgear, HVDC systems), and Wind. Between all of that, the company is currently responsible for roughly 25% of the world’s electricity and about 50% of U.S. electricity, running through an installed base of more than 7,000 gas turbines and 60,000 wind turbines. Big footprint. The stock has more than doubled over the past year and now trades at a trailing earnings multiple in the low to mid 30s.

Why This Matters For AI

Here’s the thesis in plain English. Every hyperscaler racing to build AI data centers eventually runs into the same wall. You simply cannot train or run models without power, and power infrastructure takes years to build, not quarters. GPUs can be ordered and shipped relatively fast. Gas turbines and grid transformers can’t. Lead times for heavy duty turbines run close to three years, and GE Vernova already has customers locking in production slots into 2030 and beyond.

That creates a dynamic a lot of growth investors just aren’t used to pricing in. This is a backlog driven business where revenue visibility stretches years into the future, instead of the one or two quarter visibility you’d get from a typical software or chip company. Total backlog stood at $163 billion last quarter, up $13 billion in a single quarter, with equipment orders up 71% year over year. That backlog alone represents more than two years of forward revenue at the current pace.

The electrification segment specifically, the part that connects data centers to the grid, booked $2.4 billion in data center related orders in a single quarter. That’s more than the entire prior year combined. And pricing on new gas turbine orders is running 10 to 20 percentage points higher on a dollar per kilowatt basis than it was just two quarters ago. That’s not a volume story. That’s a pricing power story, which usually only shows up when supply is genuinely tight.

Wall Street Can’t Quite Agree On What This Is Worth

Consensus price targets from Wall Street cluster somewhere in the $1,050 to $1,212 range, and about 29 of the 34 analysts covering the stock rate it a Buy. But at least one independent quant shop disagrees pretty hard, pegging a discounted cash flow fair value as low as $700 to $1,191. That kind of split is actually useful information. When the debate is about how fast something is happening rather than whether it’s happening at all, that usually means the market hasn’t settled on the right multiple yet.

Does GE Vernova Make It To Our LENS Index?

We track names like this through a separate, rules based portfolio we run at Baptista Research called the LENS Index. It’s a large cap portfolio that only opens a position after a thesis has been published and a specific catalyst has been named, never before the fact. Since it launched in May 2026, it’s been beating the S&P 500 Total Return while running at meaningfully lower risk than the index itself.

GE Vernova is under active coverage for the sake LENS right now. Whether that means it’s already sized as a real position, or whether it’s still waiting on one specific number to earn that spot, is genuinely worth understanding in detail. Because the answer tells you exactly when this setup is supposed to be acted on.

Why Watch List, Not Yet A Buy

GE Vernova sits on the LENS Watch List, added on June 4, 2026, with the stated rationale written at the time as “AI power demand. Q1 data center orders greater than all of 2025.” That’s a real, dated observation, not something written after the fact to look smart. But almost a month later, it’s still sitting on the watch list rather than in the active portfolio, and that’s on purpose.

LENS runs on one simple rule. A named, dated catalyst has to actually confirm before conviction turns into capital, no matter how good the surrounding story sounds. The specific trigger written down for this name is “Article plus Q2 orders above $3 billion, tied to earnings on July 22.” In plain terms, GEV only moves from watch list to active position if two things happen together: a published thesis, which this article now takes care of, and Q2 data center electrification orders clearing $3 billion when the company reports on July 22, 2026. Q1’s number was $2.4 billion. A print above $3 billion wouldn’t just be steady growth, it would be the first real confirmation that momentum is actually accelerating rather than just continuing along at the same pace.

Here’s the full watch list GE Vernova is currently sitting on, each name with its own dated trigger:

Why The Market Might Be Getting This Wrong

A trailing P/E in the low to mid 30s sounds expensive for an industrial company, sure. But that multiple is being slapped on a business that most investors are still modeling using old industrial assumptions. Mid single digit growth, thin margins, cyclical demand. That’s just not what this business looks like anymore, and the gap between the bulls and that one bearish DCF shop reflects exactly that unresolved reclassification.

Where The Growth & Margins Are Actually Coming From

The clearest catalyst sits in the gas turbine business. Combined gas turbine backlog and slot reservations stood at 100 gigawatts last quarter, with management guiding to at least 110 gigawatts by year end 2026. Only about 20% of that backlog is directly tied to data centers today. The other 80% is traditional utility and industrial demand, which actually matters here, because it means the AI power thesis isn’t even the majority driver yet. It’s just the fastest growing one.

Electrification segment adjusted EBITDA margins expanded roughly 590 basis points to almost 18% last quarter. Full year 2026 guidance now calls for revenue of $44.5 to $45.5 billion, company wide adjusted EBITDA margins of 12% to 14%, and free cash flow of $6.5 to $7.5 billion, all raised from the prior guide off the strength of Q1.

The Balance Sheet Situation

The company ended its most recent quarter with roughly $10 billion in cash and gross debt to adjusted EBITDA below 1x, which is pretty conservative for a business investing this aggressively into new capacity. It’s returned capital at the same time too, with around $4.6 billion in share buybacks over the past five or six quarters, plus a modest quarterly dividend, all while also funding a full buyout of its transformer joint venture, Prolec, which closed in February.

What Could Actually Break This Thesis

Three risks worth naming honestly here, not strawmen. First, the Wind segment is still a real cash drag, with EBIT losses guided near $400 million in 2026, and the company is currently in litigation with an offshore wind customer over a disputed turbine contract termination, a court has so far ruled against GE Vernova’s attempt to exit. Second, community and permitting pushback on new data center projects is real, the CEO himself acknowledged as much at a recent investor conference, and it briefly pressured the stock. Third, and maybe most important for the AI power thesis specifically, management hasn’t closed a single long term framework agreement with a hyperscaler yet despite months of negotiation. Orders are still coming in one at a time rather than through the big multi year commitments that would really cement the bull case.

The Risk Reward Setup

Downside here is contained by the same thing that makes the stock look expensive in the first place. A $163 billion backlog representing over two years of contracted forward revenue gives this company unusual earnings visibility for an industrial name, which limits how far estimates can realistically get cut without an actual demand collapse, and there’s no sign of that in the order book right now. The upside case also doesn’t require some crazy re-rating fantasy. It just requires the market to slowly reclassify this business from “industrial cyclical” to “structural AI infrastructure compounder,” and that kind of shift tends to happen gradually and then all at once, usually around a confirming data point rather than a single headline.

Wrapping This Up

The fundamentals here, record backlog, expanding margins, real pricing power, and a genuinely structural role in the AI power buildout, are strong. And the valuation debate is a fair, two-sided argument rather than an obvious mispricing in either direction. The one number worth actually watching is Q2 data center electrification order value, due alongside earnings on July 22, 2026. This is a 12 to 24 month thesis, not a trade, and the honest answer right now is patience with a clearly defined trigger, not chasing it.

Disclaimer: We do not hold any positions in the above stock(s). Read our full disclaimer here.