A little-known law has been quietly applied to your savings for years. (From Monetary Gold)

The AI trade has taken a heavy beating lately, and few corners have been spared. Chip stocks, memory names, and neocloud providers have all sold off hard. Intel (NASDAQ: INTC) has fallen about 55% from its 52-week high, Alphabet (NASDAQ: GOOGL) nearly 30%, and Nebius (NASDAQ: NBIS) about 50%. Even the many Magnificent Seven companies and hyperscalers funding the buildout haven’t escaped. Alphabet tumbled over 12% the day after its earnings report, as investors fixated on how much it’s spending rather than how fast its cloud business is growing.

That spending is substantial. Alphabet expects capital expenditures of $195 billion to $205 billion this year and has signaled that 2027 will be significantly higher. Meta Platforms (NASDAQ: META) plans equity offerings of more than $100 billion to help fund its own buildout.

For investors, the question hanging over these companies is whether all that money will eventually earn an adequate return.

But there’s another side to that equation. Every dollar a hyperscaler spends has to land somewhere, and much of it becomes revenue for the companies building the chips, networking gear, power equipment, cooling systems, and connectors inside AI data centers. As MarketBeat recently covered, Anthropic’s IPO filing revealed $221 billion in non-cancelable compute commitments with Amazon (NASDAQ: AMZN) and Alphabet alone. For the suppliers on the receiving end, the AI buildout isn’t a cost to justify. It’s demand that’s already showing up in their order books and cash flow.

The market hasn’t fully drawn that distinction yet. All five names below fell alongside the rest of the AI complex on Friday and now trade well below their 52-week highs. Here’s a closer look at each.

NVIDIA: The Biggest Recipient of All

No company has collected more of the AI spending boom than NVIDIA (NASDAQ: NVDA). In its fiscal second quarter of 2027, revenue more than doubled to $96.2 billion, data center sales rose 117% to $89.0 billion, and management guided revenue to about $108 billion for the fiscal third quarter.

NVIDIA is a clean example of this theme because of its business model. It designs its chips but outsources manufacturing, so it doesn’t carry the heavy factory costs its customers do. That shows up in net margins near 64% and a return on equity above 96%. The stock is up 12% this year but sits about 14% below its 52-week high following Friday’s 4% drop.

Analysts hold a Buy consensus with an average price target of $321.47, implying more than 50% upside, and fiscal third-quarter results are expected around Nov. 17.

From a technical perspective, bulls will want to see the stock continue to show higher-timeframe relative strength. For that to continue, NVIDIA shares will need to remain above its 200-day simple moving average (SMA), with the $190 area as the potential final line in the sand for bulls.

Vertiv: Orders Doubled, and the Stock Still Sits Well Off Its High

Vertiv (NYSE: VRT) supplies the power and thermal management equipment that keeps AI data centers running, and its third-quarter report on Oct. 28 showed just how much demand is flowing its way. Revenue jumped 61% to $4.3 billion, organic orders more than doubled, and adjusted operating margin expanded 290 basis points to 25.2%.

Management raised its full-year guidance across every key metric, lifting its sales outlook to $14.45 billion to $14.85 billion and adjusted earnings per share (EPS) to $7.10 to $7.20. Notably, adjusted free cash flow rose 50% to $652 million in the quarter, with full-year guidance raised to $2.7 billion.

Shares jumped 16% on the report, but the broader sell-off has since pulled them slightly back. The stock is still up about 45% this year, yet it sits roughly 40% below its 52-week high. Analysts hold a Moderate Buy consensus with an average target of $337.81.

While its relative outperformance has been notable over the past week, the stock still has plenty of work to do. For a downtrend break to be confirmed and momentum to swing back in the favor of the bulls, the stock will need to reclaim its 200-day SMA near $270 a share.

Amphenol: The Connector Company Riding AI

Amphenol (NYSE: APH) is one of the less obvious winners of the buildout. It makes the high-speed connectors, cables, and interconnect systems used inside AI servers and racks, and its third-quarter results, also reported Oct. 28, were exceptional. Sales rose 65% to a record $10.2 billion, adjusted EPS climbed 74% to 80 cents, and orders hit a record $12.2 billion.

The AI connection is clear in the details. IT datacom made up 48% of sales and grew 117% from a year ago, or 77% organically, and management said AI now accounts for roughly three-quarters of that business.

Free cash flow reached a record $1.6 billion, and the company raised its dividend 20% while approving a new $4 billion share repurchase program. The stock is up close to 25% this year and about 7% below its high, with a Moderate Buy consensus and an average target of $111.79.

On the chart, the stock looks well positioned. It’s spent the entire year consolidating in a wide range, and in recent days it has reclaimed several key moving averages. Like VRT, it still has work to do, though. For the bulls to feel fully in control, they will want to see the all-important resistance area, where the stock has failed from multiple times, near $88 finally give way.

Arista Networks: The Network Behind the Clusters

Arista Networks (NYSE: ANET) sells the high-performance Ethernet switches that connect thousands of AI processors inside a data center.

Like NVIDIA, it runs a relatively asset-light model, which shows in net margins above 38%. In October, Arista hit an all-time high after launching an open Ethernet rack-scale portfolio alongside Advanced Micro Devices (NASDAQ: AMD), Broadcom (NASDAQ: AVGO), Meta, and Microsoft (NASDAQ: MSFT).

The stock has since cooled sharply. It fell over 8% on Friday alone and now sits about 22% below its 52-week high, though it remains up nearly 30% this year. Arista reports third-quarter results on Tuesday, Nov. 3, which should offer a timely read on whether hyperscaler networking demand is holding up. Analysts hold a Buy consensus with an average price target of $233.20, implying about 38% upside.

Eaton: Powering the Data Center From Grid to Chip

Eaton (NYSE: ETN) rounds out the list on the electrical side. While Vertiv focuses on power and cooling inside the facility, Eaton makes the switchgear, transformers, and power distribution equipment that move electricity from the grid into the data center. It has also been leaning further into the theme.

In March, Eaton closed its nearly $9.5 billion acquisition of Boyd Thermal, adding liquid cooling technology to create what it calls a grid-to-chip offering for data center customers.

Eaton is a well-diversified business, with aerospace and vehicle segments alongside electrical, which helps explain its steadier profile, including a dividend yield of 1.2%. Shares are up nearly 16% this year, and while the stock is below its 200-day SMA, it remains above higher-timeframe support near $360 a share.

Like Arista, Eaton reports third-quarter results on Tuesday, Nov. 3.

Ahead of the print, analysts hold a Moderate Buy consensus with an average target of $484.10, implying about 30% upside.

Where the Money Actually Goes

Doubts about the spenders have largely driven the AI sell-off, and whether hyperscalers like Alphabet and Amazon will earn enough on their enormous capital budgets. For the companies on this list, that question looks different. The spending is already showing up as revenue growth, record orders, and rising cash flow. Vertiv and Amphenol both just raised their outlooks, and Arista and Eaton get their turn on Tuesday.

Of course, that doesn’t make these stocks immune. If hyperscalers eventually slow their spending, suppliers would feel it quickly, and last week’s sell-off showed how closely the market still ties the two groups together. For now, though, the money is still flowing, and these five remain among its clearest recipients.

Read this article online ›