Hey there, bargain hunter…

You can’t plug an AI data center into a power plant with vibes.

That’s the starting point.

And it’s exactly why the market is doing something less glamorous (and more profitable): it’s buying the picks-and-shovels crews that rebuild the grid—especially transmission, substations, and the big iron transformers that make the whole system work.

Scoreboard: What Actually Happened

Here are the up-to-date, non-sexy facts that matter:

• Transformer aging is not a rumor: the U.S. Department of Energy’s Large Power Transformers report pegs the average age around ~40 years, with ~70% older than 25 years. ([energy.gov](https://www.energy.gov/oe/articles/large-power-transformers-and-us-electric-grid-report-june-2012?utm_source=openai))
• Transmission planning rules got teeth: FERC Order No. 1920 (issued May 13, 2024) requires long-term (at least ~20-year) regional transmission planning and updates cost allocation. ([ferc.gov](https://www.ferc.gov/explainer-transmission-planning-and-cost-allocation-final-rule?utm_source=openai))
• Capital is moving from “ideas” to “invoices”: Quanta Services ended 2025 with record backlog of $44.0B and reported 2025 free cash flow of $1.7B. ([investors.quantaservices.com](https://investors.quantaservices.com/news-events/press-releases/detail/390/quanta-services-reports-fourth-quarter-and-full-year-2025-results/?utm_source=openai))
• Primoris is sitting on real work, not just a story: total backlog was $11.5B as of June 30, 2025 (with about $6.0B utilities backlog). ([ir.prim.com](https://ir.prim.com/news-and-events/news-releases/2025/08-04-2025-211559739?utm_source=openai))
• NERC is flagging load growth pressure: its 2025 Summer Reliability Assessment forecast aggregated peak demand rising by ~10 GW across all 23 assessment areas—more than double the 2023→2024 increase. ([publicpower.org](https://www.publicpower.org/periodical/article/record-load-growth-high-temperatures-expected-strain-grid-summer-nerc?utm_source=openai))

The Real Reason This Is Happening (Expectations vs. Reality)

The market spent years pricing the grid like it was “good enough.”

Reality check: electrification + data centers + reliability politics = the “good enough” era is over.

Transmission is the bottleneck because:

• Generation can be built in chunks.
• Data centers can be built in chunks.
• Transmission upgrades are slow (permits, siting, interconnection, transformer supply, specialized labor).

So capital is flowing to the companies that can actually execute: build lines, rebuild substations, replace transformers, and do it at scale.

Deep Dive: What the “Grid Contractor” Business Really Is

Think of grid contractors as the “construction + engineering operating system” for utilities and grid operators.

They make money by:

• Engineering + construction services for high-voltage transmission lines (the long-distance highways for electrons).
• Substation work (where voltage gets stepped up/down and power gets routed).
• Maintenance + storm response (recurring, urgent, often high-ROI work for utilities).
• Program management for multi-year capex plans (sticky customer relationships).

This isn’t a “one quarter” trade. It’s a multi-year capex cycle. And it’s politically protected because reliability failures are not tolerated for long.

Data Section: The Numbers That Matter (Backlog, Cash, and Visibility)

Here’s the Cheap Investor way to look at this theme: backlog is your demand signal, and cash flow is your lie detector.

1) Quanta Services (PWR): the backlog king

• Backlog: $44.0B at year-end 2025 (record). ([investors.quantaservices.com](https://investors.quantaservices.com/news-events/press-releases/detail/390/quanta-services-reports-fourth-quarter-and-full-year-2025-results/?utm_source=openai))
• Cash generation: $1.7B free cash flow in 2025 (record). ([sa.marketscreener.com](https://sa.marketscreener.com/news/quanta-services-4q25-and-full-year-2025-operational-financial-commentary-ce7e5ddcd88efe25?utm_source=openai))
• Electric segment backlog: cited at $36.2B at year-end 2025 (record), showing the “grid” engine is doing the heavy lifting. ([sa.marketscreener.com](https://sa.marketscreener.com/news/quanta-services-4q25-and-full-year-2025-operational-financial-commentary-ce7e5ddcd88efe25?utm_source=openai))

Translation: PWR is already the obvious name. It’s the “quality compounder” people default to when they want grid exposure.

2) Primoris Services (PRIM): smaller, less crowded, real utilities backlog

• Total backlog: $11.5B as of June 30, 2025. ([ir.prim.com](https://ir.prim.com/news-and-events/news-releases/2025/08-04-2025-211559739?utm_source=openai))
• Utilities backlog: about $6.0B as of June 30, 2025. ([ir.prim.com](https://ir.prim.com/news-and-events/news-releases/2025/08-04-2025-211559739?utm_source=openai))

Translation: PRIM gives you grid exposure with a different investor base than PWR—sometimes that matters when “the obvious trade” gets crowded.

3) MYR Group (MYRG): pure-play flavor with T&D exposure

• T&D segment backlog: reported around $927M (last twelve months ended June 30, 2025, per company coverage/slides). ([investing.com](https://www.investing.com/news/company-news/myr-group-q2-2025-slides-strong-performance-amid-data-center-clean-energy-boom-4288463?utm_source=openai))

Translation: MYRG is often more sensitive to project timing and execution, but it can work as the “spicier” contractor bet when the cycle accelerates.

Is It Cheap? (The Cheap Investor Section You Asked For)

“Cheap” doesn’t mean low P/E. Cheap means you’re not paying for perfection.

Right now, the grid contractor bucket splits into two camps:

• Not cheap (often “quality priced”): PWR tends to trade like the category leader (because it is). You’re paying for execution + visibility.
• Potentially cheaper (less loved / less institutionally crowded): smaller names like PRIM and MYRG can sometimes offer better entry points when sentiment swings or quarterly timing gets messy.

My framework for “cheap” here is simple:

• Do you have multi-year backlog?
• Do you convert that backlog into free cash flow?
• Do you avoid balance-sheet stress while you scale?

PWR checks those boxes cleanly on backlog + cash generation. ([investors.quantaservices.com](https://investors.quantaservices.com/news-events/press-releases/detail/390/quanta-services-reports-fourth-quarter-and-full-year-2025-results/?utm_source=openai))

The “cheap investor” opportunity is usually not buying the best business at any price. It’s buying a solid business when the market is temporarily bored—and this theme is boring by default.

Bull case

Load growth forces accelerated transmission + substation spend. Order 1920-driven planning turns into real project awards, and contractors sustain pricing power in labor and equipment.

Base case

Spending ramps, but lumpy permitting and supply chain timing create “feast/famine” quarters. Backlog stays high; returns are solid but not linear.

Bear case

Projects get delayed (siting fights, cost allocation disputes), margins compress from labor/material inflation, and the market re-rates contractors down even if long-run demand is intact.

Action Plan (Cheap Investor Style)

Don’t overcomplicate it, bargain hunter.

• If you want the “core” exposure: build a starter position in the category leader (PWR) and add on pullbacks tied to timing noise—not thesis breaks.
• If you want “cheaper beta” to the theme: watch PRIM and MYRG for sentiment dips (guidance conservatism, weather delays, project timing) while backlog stays intact.
• Scale-in framework: 3 buys over 6–12 weeks. Your edge is patience, not prediction.

Cheap Investor Checklist / Scorecard (Track These)

• PWR backlog trend: does it stay near record levels after 2025’s $44.0B? ([investors.quantaservices.com](https://investors.quantaservices.com/news-events/press-releases/detail/390/quanta-services-reports-fourth-quarter-and-full-year-2025-results/?utm_source=openai))
• PWR free cash flow: does cash conversion stay strong after $1.7B FCF in 2025? ([sa.marketscreener.com](https://sa.marketscreener.com/news/quanta-services-4q25-and-full-year-2025-operational-financial-commentary-ce7e5ddcd88efe25?utm_source=openai))
• PRIM backlog stability: does utilities backlog remain around the $6B neighborhood? ([ir.prim.com](https://ir.prim.com/news-and-events/news-releases/2025/08-04-2025-211559739?utm_source=openai))
• NERC load growth updates: do forecasts keep stepping up (e.g., the ~10 GW summer peak demand increase signal)? ([publicpower.org](https://www.publicpower.org/periodical/article/record-load-growth-high-temperatures-expected-strain-grid-summer-nerc?utm_source=openai))
• Transformer constraints: do aging stats + supply constraints remain a policy priority (average LPT age ~40 years)? ([energy.gov](https://www.energy.gov/oe/articles/large-power-transformers-and-us-electric-grid-report-june-2012?utm_source=openai))
• Order 1920 implementation: do regions actually file compliant long-term plans and cost allocation frameworks, or does it bog down in process? ([ferc.gov](https://www.ferc.gov/explainer-transmission-planning-and-cost-allocation-final-rule?utm_source=openai))
• Margin discipline: are contractors holding margins as labor costs rise?

Bottom Line

If AI load keeps rising and electrification stays a policy priority, then the grid has to be rebuilt—slowly, expensively, and in public.

Most investors want the shiny part (chips, data centers, generation). The money, frustratingly often, ends up with the crew doing the wiring.

If you want one clean signal that the theme is real: follow backlog and free cash flow. The rest is noise.