Harnisch Put 91% of Peconic’s $4.2B Into Three Infrastructure Stocks After Returning 79% in 2025: The Power Grid Play That Crushed the S&P 500

A 57-year-old hedge fund manager who started on Wall Street in 1968 just turned a $140 million portfolio into a $4.2 billion concentrated bet on three companies most retail investors have never heard of. Bill Harnisch's Peconic Partners delivered 79% returns in 2025 — crushing the S&P 500 for the fifth time in six years — and his Q4 2025 13F filing reveals exactly where he's putting his conviction: 91.5% of the entire fund in Quanta Services, Dycom Industries, and MasTec.

These aren't AI chip makers or cloud hyperscalers. They're the companies that physically build the power lines, fiber networks, and electrical infrastructure that AI actually needs to run. While the market debates GPU margins and LLM pricing, Harnisch is betting billions that the real bottleneck — and the real opportunity — is in the ground.

TL;DR — Peconic Partners Q4 2025

• AUM exploded from $140M to $4.21B — a +2,907% increase in a single quarter
• 91.5% of the portfolio sits in just three stocks: Quanta Services (PWR), Dycom Industries (DY), and MasTec (MTZ)
• Quanta Services alone is half the fund at $2.1B and 50.2% of AUM
• Complete rotation out of solar tracker stocks — sold Shoals Technologies (SHLS) and Array Technologies (ARRY) entirely
• Rotated into power grid and fiber infrastructure builders — the physical picks-and-shovels of the AI electricity boom
• 79% full-year 2025 returns per Bloomberg, outperforming the S&P 500 for the fifth time in six years
• Natural gas positions retained from Q3 — EQT Corp and Range Resources supply fuel for new power generation
• Freeport-McMoRan (FCX) added as a copper play on electrification demand
• Portfolio narrowed from 6 to 15 holdings but concentration increased massively, with top-3 at 91.5%
• Harnisch's thesis: AI requires enormous electricity and data infrastructure — the builders, not the chip makers, are the overlooked winners

The Three Stocks That Built a 79% Year

The headline numbers are staggering. Peconic's top three positions account for $3.85 billion of the $4.21 billion portfolio. Each of these companies does roughly the same thing: they build and maintain the physical infrastructure that carries electricity and data across America.

Quanta Services (PWR) — $2,114M | 50.2% of Portfolio

Half of Peconic's entire fund sits in Quanta Services. The Houston-based company is the largest specialty contractor in North America for electric power, oil and gas, and telecommunications infrastructure. When utilities need high-voltage transmission lines built, when data centers need grid connections, when broadband fiber needs to reach rural communities — Quanta does the work.

This isn't a speculative position. Quanta has a $30+ billion backlog driven by grid modernization, renewable energy interconnection, and the explosive growth in data center power demand. Every new AI training cluster needs grid capacity that didn't exist two years ago, and Quanta is one of the few companies with the workforce and expertise to build it.

Dycom Industries (DY) — $1,388M | 33.0% of Portfolio

At one-third of the portfolio, Dycom is Peconic's second-largest conviction bet. Dycom is the leading specialty contractor for telecommunications providers — they install and maintain the fiber optic cables, underground conduit, and aerial lines that carry internet traffic. Major customers include AT&T, Lumen Technologies, and Comcast.

The fiber buildout story is multi-year. The federal Broadband Equity, Access, and Deployment (BEAD) program alone is distributing $42.5 billion to expand high-speed internet access. Dycom is a primary beneficiary. The stock surged 26% in November 2025 alone as contract wins accelerated.

MasTec (MTZ) — $348M | 8.3% of Portfolio

MasTec completes Harnisch's infrastructure trifecta. The Miami-based contractor operates across power delivery, clean energy, oil and gas pipelines, and telecommunications. Where Quanta specializes in electrical and Dycom in fiber, MasTec spans both — plus pipeline construction for natural gas delivery to new power plants.

Together, these three companies represent the physical backbone of America's energy transition and digital infrastructure buildout. Harnisch isn't betting on which AI model wins — he's betting that all of them need more power and more bandwidth.

The Solar-to-Infrastructure Rotation

What makes Peconic's Q4 filing so striking isn't just what Harnisch bought — it's what he sold. In Q3 2025, the portfolio was a modest $140 million split across six positions, with solar tracker companies Shoals Technologies (SHLS) and Array Technologies (ARRY) as the two largest holdings at a combined 50.6% of the fund.

By Q4, both were gone entirely. Harnisch didn't just trim them — he liquidated and redeployed at massive scale.

The logic is revealing. Solar trackers (the mechanical systems that angle solar panels toward the sun) are a competitive, commoditizing market. Shoals has faced margin pressure and product quality concerns. Array has seen pricing competition from Chinese manufacturers. The solar tracker thesis was playing out, but the returns were compressing.

The infrastructure builders, by contrast, face limited competition. You can't offshore the construction of a 500kV transmission line across Texas. The skilled labor shortage in electrical line work creates a natural moat. And the demand pipeline — from grid modernization to data center connections to BEAD broadband — stretches years into the future.

Harnisch essentially moved up the value chain: from solar component manufacturing to the companies that install everything.

The Energy Supply Chain — Supporting Positions

Beyond the big three, Peconic's smaller positions tell a coherent story. Every holding connects back to the same thesis: America needs dramatically more energy infrastructure.

Freeport-McMoRan (FCX) — $102M (2.4%)
New position. The world's largest publicly traded copper producer. Electrification requires copper — for wiring, transformers, EV charging stations, and grid connections. Copper demand from AI data centers alone is projected to add 1-2% to global demand annually. This is Harnisch playing the same infrastructure thesis through commodities.

First Solar (FSLR) — $77M (1.8%)
After selling solar tracker companies SHLS and ARRY, Harnisch kept exposure to solar through First Solar — a domestic panel manufacturer that benefits from IRA tax credits and tariff protection. It's a more defensible position in the solar value chain than the trackers he exited.

Amazon (AMZN) — $70M (1.7%)
A small position in the context of a $4.2B fund. Interesting because Harnisch is known to short retailers. Amazon's presence here likely reflects data center and AWS demand as a driver of the infrastructure buildout Peconic is positioned to benefit from. The demand signal, not the stock itself, is the point.

EQT Corp (EQT) — $31M (0.7%) and Range Resources (RRC) — $28M (0.7%)
Both held over from Q3. Appalachian natural gas producers that supply fuel for gas-fired power plants — the dispatchable generation that backs up renewables and powers data centers around the clock. Natural gas is the bridge fuel for AI's electricity demands, and these positions anchor the energy supply side of Harnisch's thesis.

What Harnisch Sees That Others Miss

Bill Harnisch has been investing for 57 years. He's seen cycles, bubbles, and paradigm shifts. His current positioning suggests he sees three overlapping mega-trends converging on the same set of companies:

1. AI Electricity Demand Is Just Starting
As Harnisch told Bloomberg: "We haven't seen how much AI can contribute, but we know one thing: It's going to be a lot more data. It's going to require a lot more electricity." Current estimates suggest AI data centers could consume 3-5% of U.S. electricity by 2030, up from roughly 2% today. That incremental demand requires massive grid expansion — new transmission lines, new substations, new interconnections. Quanta builds all of them.

2. Broadband Infrastructure Is Federally Funded
The BEAD program's $42.5 billion is just beginning to flow into actual construction contracts. Dycom is positioned as a primary contractor. Unlike most government spending programs, this money is earmarked for physical infrastructure that must be built — it can't be redirected to software or services.

3. Clean Energy Transition Requires Grid Modernization
Renewable energy projects (solar farms, wind farms, battery storage) are only as good as the transmission lines connecting them to the grid. The U.S. has a well-documented transmission bottleneck — interconnection queues have ballooned to over 2,600 GW of capacity waiting to connect. Every project that clears the queue needs a company like Quanta or MasTec to build the physical link.

Harnisch's macro uncertainty is itself part of the thesis. He told Bloomberg: "If you asked me where GDP is going to be next year, I could paint with ten different stories. I have no idea." But he doesn't need to know GDP direction. Whether the economy booms or stalls, data centers still need power, broadband still needs fiber, and the grid still needs upgrading. The demand for infrastructure is structurally inelastic.

What Analysts Might Misread

From the outside, a 91.5% concentration in three mid-cap construction companies looks reckless. But context matters.

These aren't momentum trades. PWR, DY, and MTZ don't show up on most retail screeners or AI stock lists. They're classified as "construction & engineering" or "specialty contractors" — sectors that institutional growth investors typically ignore. That's precisely the point. Harnisch is buying where the smart-money consensus isn't yet crowded.

The 13F only shows the long book. Peconic is a long/short fund. Harnisch is known to short retailers like Walmart, tech companies like Amazon (in other contexts), and industrials like GE Vernova and Eaton. The 91.5% concentration in infrastructure longs is paired with an unknown short book that may hedge macro risk, sector rotation, or specific company exposure. The long-only 13F view overstates the portfolio's directional risk.

The AUM jump isn't all new capital. Going from $140M to $4.2B in one quarter could reflect new investor inflows (driven by the 79% return headline), strategy changes in what's reported on 13F, or shifts between asset classes. The conviction shown in the position sizing is real regardless — Harnisch chose to put 91% of whatever he deployed into three names.

Frequently Asked Questions

Why did Peconic's AUM jump from $140M to $4.2B in one quarter?

The most likely explanation is a combination of strong performance attracting new investor capital (79% annual returns are a powerful marketing tool), plus potential changes in how the fund deploys assets across reportable and non-reportable strategies. The 13F only captures U.S.-listed equity holdings, so shifts from non-13F assets (options, private investments, non-U.S. stocks) into U.S. equities would also show as a jump.

Is 91% concentration in three stocks normal for a hedge fund?

It's extremely unusual. Most institutional portfolios hold 30-100+ positions with no single stock above 10-15%. Peconic's concentration is closer to a concentrated activist fund or a private equity approach. However, Harnisch has a 57-year track record and explicitly favors high-conviction bets. The long/short structure also means the net exposure is less extreme than the 13F suggests.

What do Quanta Services, Dycom, and MasTec actually do?

All three are specialty infrastructure contractors. Quanta builds and maintains electric power transmission and distribution systems. Dycom installs fiber optic and telecommunications infrastructure. MasTec spans power delivery, clean energy construction, oil & gas pipelines, and telecom. Together, they build the physical systems that carry electricity and data across North America.

Why did Harnisch sell his solar tracker positions (SHLS, ARRY)?

Solar trackers have become increasingly commoditized with growing competition from Chinese manufacturers. Shoals Technologies faced product quality concerns and margin pressure. Rather than bet on individual solar components, Harnisch moved into the companies that install solar and other energy projects — a more defensible position in the clean energy value chain with higher barriers to entry.

How does this portfolio relate to the AI investment theme?

Harnisch is playing AI through its physical requirements rather than its software. AI training and inference require massive amounts of electricity and data bandwidth. Quanta builds the grid connections for data centers. Dycom lays the fiber that carries data. MasTec builds the power plants and pipelines. It's the picks-and-shovels approach applied to AI's energy bottleneck — a thesis that works regardless of which AI company ultimately wins.

Where can I track Peconic Partners' full holdings?

You can view Peconic Partners' complete 13F filing history, position changes, and portfolio analysis on their 13F Insight filer page. The page includes quarter-over-quarter comparisons, sector breakdowns, and historical AUM trends.