An Energy Crisis is Brewing

In partnership with

Happy Sunday

Today I want to talk about a theme I think is in the 3rd inning. Electricity. The Grid. Energy.

We don't have enough, and that creates an opportunity.

Most folks don't know this, but I've written for two energy publications, my father held his solar contracting license, and I'm a lifelong surfer and intracoastal resident who cares deeply about energy and our environment.

Because of all that, I keep myself updated on energy and utility trends. I'm seeing a clear market inefficiency unfold.

As always, this isn't personal financial advice. It's market commentary from my perspective.

Here's why I think there's something real here

AI and data centers and the need for computation is breaking the US power grid

OpenAI, Amazon, Google, and Microsoft’s AI tools use 10-20x more electricity than normal computing, and demand jumped 3% last year after being flat for 40 years. Microsoft is cancelling billion-dollar datacenter projects because power companies can't give them electricity.

The equipment that moves electricity around is on back order: 3-6 years to build 

Combined transformer backlog exceeds $119 billion through 2028. Transformers…the equipment that converts and moves high-voltage electricity, now require 3-6 year lead times versus 1 year historically. The US produces almost none of the specialized steel required and relies on a global oligopoly of three major manufacturers. Combined backlog exceeds $119 billion through 2028. Manufacturing capacity cannot scale quickly because each transformer is custom-engineered for specific voltage and load requirements, with components that require lengthy procurement and precision manufacturing.

The obvious stocks ran 150-250%, but that's a rotation signal not an exit signal

The market figured this out. Equipment manufacturers like GE Vernova, Vertiv, and Eaton already moved up 150-250% in 12 months and now trade at 150x+ P/E ratios. At those prices, you need perfect execution for years just to justify today's valuation.

Stretched valuations on the leaders don't mean the theme is wrong. Transformer backlogs still run through 2028. The bottleneck is still real. What it means is the market priced the manufacturers but ignored their suppliers.

When the heroes get overextended, the opportunity rotates downstream. The same 3+ year infrastructure bottleneck that drove GE Vernova to 150x P/E creates demand for electrical steel, copper, and natural gas—and those suppliers trade at 6-17x P/E because the energy crunch hasn't connected the dots yet.

This runs for 2-3 more years minimum because you can't fix it faster

Even if you wanted to solve the shortage tomorrow, physics won't let you manufacture transformers faster. New capacity doesn't come online until 2027. The bottleneck has 2-3 years left to run.

I'm focused on two groups the market is mispricing:

Natural gas pipelines - Williams Companies, Kinder Morgan, Energy Transfer trade at 15-30x P/E with 4-7% dividends. Most folks lump them in with oil stocks, but every datacenter runs on natural gas backups and peaker plants. Gas plants take 12-24 months to build versus 5-10 years for nuclear. If datacenters actually need all this power, they're burning more natural gas. S&P estimates 3-6 billion cubic feet per day more by 2030. These stocks have been consolidating for a year.

Steel and copper - Cleveland-Cliffs makes the only US supply of electrical steel that goes into transformers. You can't build one without it. Trades a negative P/E (they are currently losing money). Copper miners like Freeport-McMoRan supply 7-15 tons per transformer. If GE Vernova has $119 billion in transformer orders through 2028, somebody's gotta supply the raw materials. They trade at 1/10th the price because people buy the finished product, not the components. I'm interested in the supply chain of these transformers.

The Setup

This works if datacenter demand is real and not overhyped like 1999. If it's phantom demand or microgrids bypass utilities entirely, the trade loses. But the risk-reward at 6-17x P/E looks a lot better than chasing manufacturers at 150x—and gas pipelines pay 4-7% dividends while you wait.

The physical bottleneck gives this 2-3 years to run. That's the window.

-John

Today’s Sponsor

The Free Newsletter Fintech Execs Actually Read

If you work in fintech or finance, you already have too many tabs open and not enough time.

Fintech Takes is the free newsletter senior leaders actually read. Each week, we break down the trends, deals, and regulatory moves shaping the industry — and explain why they matter — in plain English.

No filler, no PR spin, and no “insights” you already saw on LinkedIn eight times this week. Just clear analysis and the occasional bad joke to make it go down easier.

Get context you can actually use. Subscribe free and see what’s coming before everyone else.

Sign Up Free