Edited by humans. Written by AI. How our editing works
All articles

4 Robotics Stocks Quietly Building the Next Wave

ARM, Symbotic, Teradyne, and BOTZ are positioned at the heart of physical AI. Here's what the robotics investment thesis actually looks like up close.

Rachel "Rach" Kovacs

Written by AI. Rachel "Rach" Kovacs

May 21, 20267 min read
Share:
Man wearing glasses smiling at camera with purple tech background and yellow text reading "THESE WILL BE WORTH TRILLIONS

Photo: AI. Soraya Hadid

The Nvidia trade is over. Not because Nvidia is done — they're not — but because everyone already knows about it. The 1,400% five-year run is now a story people tell at dinner parties, not a thesis they're quietly building a position around.

So where does the next wave come from? A recent breakdown from the Fin Tek channel on YouTube makes a case worth examining: the robotics buildout, specifically what they call "physical AI" — machines that don't just process language but navigate the physical world in real time. Four companies get the spotlight: ARM Holdings (NASDAQ: ARM), Symbotic (NASDAQ: SYM), Teradyne (NASDAQ: TER), and the Global X Robotics and Artificial Intelligence ETF (NASDAQ: BOTZ).

The argument isn't that these companies will definitely outperform Nvidia. It's that they haven't already run up the way Nvidia has, and that they sit at chokepoints in a market projected to hit $110 billion annually by 2030 — not counting autonomous vehicles.

That's a structural argument, not a momentum one. And structural arguments are worth taking seriously, even when they're packaged inside a YouTube video with affiliate links.


ARM Holdings: The Unsexy Infrastructure Play

ARM's pitch is almost counterintuitive. The company behind the chip architecture powering your phone — and a growing share of industrial and edge computing devices — is currently valued at roughly $240 billion. That sounds big until you remember Nvidia is sitting north of $2.5 trillion. ARM is less than 10% of that.

The Fin Tek analysis frames ARM's advantage around a specific architectural decision in physical AI: where does the intelligence actually run? Cloud-based AI can be powerful but introduces latency. For a robot sorting packages or navigating a factory floor, a half-second delay isn't a minor inconvenience — it's a failure mode. Edge computing, where processing happens on the device itself, solves that latency problem, but it demands chips that deliver real performance without overheating or draining power fast.

That's ARM's home turf. Low-power, high-efficiency mobile chip design is what they've been optimizing for decades.

What's newer — and more interesting for investors — is ARM's move up the value stack. Rather than just licensing their architecture to others, they're now designing chips outright. Their AGI CPU, built for AI data centers, is designed to complement Nvidia's GPUs rather than compete with them. As ARM's executive vice president for cloud AI put it: "Customers like AWS and Google and Microsoft and Nvidia have all come out and said, 'Hey, this is a good thing for the ecosystem.'" At least eight customers have already committed to purchasing it.

The honest risk here is timing. ARM has stated the AGI CPU won't meaningfully impact earnings until March 2028. By standard valuation metrics, the stock is expensive today. The Fin Tek take: you don't wait until 2028 to buy in, because by then it's priced in. That logic is defensible — but it requires patience and tolerance for a stock that may not reward you for two or three years.


Symbotic: When the Market Misreads Good News

The Symbotic situation is genuinely strange, and the strangeness is the point.

The warehouse automation company recently reported its first-ever profitable quarter. The stock dropped 20%. That kind of disconnect — good news, bad price reaction — usually means one of two things: either the market knows something the headline doesn't, or it's a mispricing.

The Fin Tek case is that it's the latter. Symbotic has $22 billion in contracted backlog, a freshly profitable income statement, a massive Walmart deal (they acquired Walmart's entire in-house automation division in 2025), and a new contract with the largest co-op food wholesaler in the country, with deployment expected in early 2027.

The context for why this market matters: third-party logistics represents roughly 10% of global GDP. Every efficiency gain in that system is economically significant at scale. Amazon has already crossed one million warehouse robots and is adding more robots than human employees annually. The demand is real.

But Symbotic's risks are real too, and the Fin Tek analysis doesn't paper over them. The company is heavily concentrated in Walmart as a customer. Their stock rose 300% in 2025 before this pullback, so the current drop is partly just gravity. And the central question the market is actually asking isn't whether robotics works — it's whether Symbotic can execute on its existing pipeline at the margins their valuation requires.

"The demand side looks real. The real question is operational execution," the analysis notes. That's the cleanest articulation of the bull/bear split on this stock. Leadership credentials are solid — a CEO with supply chain roots, a tech lead from Google's robotics division via Stanford — but credentials don't ship deployments on time.

For investors who can absorb concentration risk and a potentially lumpy execution timeline, the current valuation (post-correction) may represent a genuine entry point. For those who can't, BOTZ — the ETF — offers Symbotic exposure without single-stock dependency.


Teradyne: The Quiet Infrastructure Layer

The most underappreciated pick in the Fin Tek breakdown might be Teradyne, and the underappreciation is partly by design.

Teradyne doesn't make the flashy robots. They make the automated testing systems that verify whether the chips inside those robots actually work. As robotics deployments scale and chip complexity increases, the demand for that testing infrastructure grows with it. The automated test equipment market currently sits above $10 billion annually and is expected to more than double by 2032.

More to the point: it's essentially a duopoly. Teradyne and Advantest dominate the space. Duopolies have pricing power. Pricing power matters.

The second business line is cobots — collaborative robots, branded Universal Robots, designed to work alongside humans in industrial settings. These aren't the sci-fi robots that live in cleanrooms and can't be touched. They're deployed in warehouses and fabrication shops, flexible enough to shift between tasks. Teradyne has been partnering with Nvidia's simulation platform to develop this side of the business.

The reshoring manufacturing trend is a potential accelerant here. If more fabrication work moves back to the US or nearby countries — a trend that's been building for a few years now — the demand for industrial automation in those facilities is a logical downstream effect.

The bold AI investment bets large tech companies are currently making also matter for Teradyne in a less obvious way: more AI hardware investment means more chips being produced, which means more chips requiring testing.


BOTZ: The Hedge for the Undecided

The fourth recommendation is structural rather than company-specific. The Global X Robotics and Artificial Intelligence ETF (BOTZ) holds stakes across the robotics landscape including medical robotics, international players, and industrial automation — categories that the other three picks don't cover.

For someone persuaded by the robotics thesis but less certain about which specific companies execute best, BOTZ distributes that uncertainty across the sector. The tradeoff is upside: an ETF that includes the winners also includes the companies that stumble, which caps returns relative to a concentrated position in the right single stocks.

The Fin Tek presenter mentions considering BOTZ personally for exactly this reason — to spread exposure rather than bet on execution at any one company. That's an honest framing, and it reflects a real tension that every investor in an emerging sector faces.


There's a version of this robotics thesis that's pure hype — "physical AI" is a term that's easy to attach to almost anything mechanical. But the companies here are generating real revenue, winning real contracts, and sitting at genuine chokepoints in the supply chain for automation technology.

The question isn't whether the robotics market grows. Most credible projections point the same direction. The question is whether these specific companies can translate that market growth into the earnings that justify their current valuations — and whether that happens on a timeline investors can actually hold through.

That's the bet. The terms are transparent. Whether those terms work for you depends entirely on your own risk tolerance and time horizon — not on whether the robots are real.


Rachel "Rach" Kovacs is Buzzrag's cybersecurity and privacy correspondent.

From the BuzzRAG Team

We Watch Tech YouTube So You Don't Have To

Get the week's best tech insights, summarized and delivered to your inbox. No fluff, no spam.

Weekly digestNo spamUnsubscribe anytime

More Like This

Bearded man wearing glasses and blue beanie gestures toward castle display, with text overlay about Microsoft losing $350…

AI Investments: Bold Moves and Uncertain Outcomes

Meta and Microsoft face diverging market reactions to AI investments, revealing deeper industry tensions.

Bob Reynolds·6 months ago·3 min read
A curved line graph with a cute smiling face at the center, a red lobster on the left, and labels for "ASTRAL" and "TBPN"…

OpenAI and Anthropic's Developer Tool Land Grab

OpenAI acquired Astral. Anthropic acquired Bun. Developers cheered one and panicked at the other. Here's what the different reactions actually reveal.

Marcus Chen-Ramirez·2 months ago·7 min read
A man in a business suit presents alongside a humanoid robot with pink hair and pale synthetic skin against a tech backdrop…

Humanoid Robots in 2026: Softer, Smarter, Scarier

From warm-skinned social companions to armed robotic wolf packs, humanoid robots in 2026 are forcing questions we don't have answers to yet.

Yuki Okonkwo·2 months ago·7 min read
Older man in business attire with concerned expression next to a stock chart showing sharp decline in 2026, with text "THIS…

Wall Street Knew the Crash Was Coming. Saying So Got You Fired.

Jeremy Grantham's famous poll revealed 398 analysts knew the dot-com crash was guaranteed. A 2003 study shows why none of them said so publicly: accuracy cost careers.

Carmen Rodriguez·2 months ago·8 min read
Man in blue shirt smiling at camera with "$125M Growth Engine" text and upward arrow graphic on blue background, GoFundMe…

GoFundMe's CPTO on Marketplace Failure & AI Growth

GoFundMe's CPTO Arnie Katz breaks down 3 marketplace failure modes and how AI agents are driving real revenue—not just dev productivity.

Yuki Okonkwo·2 months ago·8 min read
Bloomberg Investigates title screen with grid of pixelated video footage and faces on dark background, Channel 3 logo visible

Russia's Disinformation Playbook Hits Main Street

Storm-1516's AI deepfake operation isn't just geopolitics—it's the same smear playbook that's been destroying small businesses for years. Here's what you need to know.

Dorothy "Dot" Williams·2 months ago·8 min read
Developer at desk with dual monitors displaying analytics dashboards and data visualizations in a neon-lit tech environment

34 Self-Hosted Projects That Could Replace Your Cloud Stack

From AI email agents to thermal printer dashboards, these trending GitHub projects show what happens when developers get tired of subscription fees.

Rachel "Rach" Kovacs·3 months ago·6 min read
Bearded man wearing glasses and a beanie gestures toward camera with confused expression, text reads "NOW WHAT?

Why Your AI Agent Sits Idle After Installation

Installing an AI agent takes 10 minutes. Making it actually useful takes 40 hours. Here's why the industry keeps solving the wrong problem.

Rachel "Rach" Kovacs·3 months ago·6 min read

RAG·vector embedding

2026-05-21
1,851 tokens1536-dimmodel text-embedding-3-small

This article is indexed as a 1536-dimensional vector for semantic retrieval. Crawlers that parse structured data can use the embedded payload below.