Arm Holdings plc stock is having one of those “welcome to modern markets” weeks: the long-term AI narrative is still alive and well, but short-term sentiment has turned skittish as analysts debate valuation, Arm’s next strategic move, and what parts of the AI boom Arm actually captures.
As of the latest available quote early Thursday, Dec. 18, 2025, Arm Holdings plc (NASDAQ: ARM) traded around $114.58, down about 5.4% from the prior close.
Below is what’s driving the move, what Wall Street is forecasting now, and the big catalysts (and landmines) investors are watching next.
Why Arm stock is moving now: downgrades collide with a jittery AI tape
Goldman Sachs turns bearish on ARM
A key near-term headline: Goldman Sachs downgraded Arm to “Sell” from “Neutral”, arguing the company has limited leverage to the AI cycle compared with more direct AI hardware plays. In a note circulated via Reuters/Refinitiv coverage, Goldman also flagged that Arm’s expected higher R&D spending—especially tied to ambitions beyond pure IP licensing—could reduce financial leverage in coming fiscal years. TradingView+1
That thesis matters because Arm is widely held as an “AI infrastructure” proxy. When a major bank says, in effect, “the AI excitement may not flow into fundamentals as cleanly here,” it can be enough to knock the stock—particularly when valuations are already doing parkour.
The broader AI trade has been volatile
Arm’s pullback isn’t happening in a vacuum. Recent market action has shown investors getting more selective about AI spending and returns. For example, Reuters reported that disappointing forecasts and elevated spending concerns around Oracle helped spark a tech selloff that pulled down multiple AI-related stocks, including Arm.
This is the recurring 2025 pattern: markets love AI, then they interrogate the bill.
Geopolitics just poured gasoline on the semiconductor narrative
Adding fresh spice to an already spicy sector: Reuters reported details of China’s state-backed push toward EUV lithography capability—an effort meant to reduce reliance on Western-controlled advanced chipmaking tools. While this is more about manufacturing than CPU instruction sets, it reinforces how strategic (and politically sensitive) the entire semiconductor stack has become.
For investors, that kind of story tends to raise two opposing possibilities at once:
- more urgency (and capital) across chips and AI infrastructure, and
- more geopolitical risk premiums, export controls, and policy surprises.
Arm sits downstream of both.
Arm’s fundamentals: what the company just delivered
Despite recent stock weakness, Arm’s most recent reported quarter showed strong year-over-year growth.
Arm said its second-quarter fiscal 2026 revenue (quarter ended Sept. 30, 2025) was $1.14 billion, up 34% year over year, with royalty revenue up 21% to $620 million and licensing revenue up 56% to $515 million.
That mix is important:
- Royalties are the “flywheel” business: Arm gets paid per chip shipped using its designs.
- Licensing can be lumpier (big contracts timing), but it’s also where new platforms and broader system-level offerings can show up.
Arm’s CEO Rene Haas has repeatedly framed the data center opportunity around a brutally simple constraint: power. In Reuters’ coverage of the results, Haas pointed to AI compute demand and energy efficiency as a structural tailwind for Arm’s architecture in servers.
Arm also guided for a fiscal third quarter revenue midpoint that topped analyst expectations at the time, according to Reuters.
The bull case for Arm stock: Armv9 royalties, data centers, and “AI everywhere”
Arm’s optimistic thesis hasn’t gone away. It’s just being forced to show its work.
1) Armv9 and higher-royalty product cycles
Arm’s newer architecture and more complete design offerings are central to the “earn more per chip” story.
In its results commentary, Arm highlighted the ongoing adoption of Armv9 and Arm Compute Subsystems (CSS) as drivers of higher royalty rates per chip.
Reuters also noted that Arm’s CSS is a more complete chip design approach that helps customers build full chips faster—and that it can generate higher royalties than some older approaches.
2) Data center penetration: Arm wants a real slice of server CPUs
Arm has long dominated smartphones, but the bigger prize is the data center: higher ASPs, long upgrade cycles, and AI workloads that punish inefficient compute.
Reuters reported that Arm has said it expects its share of CPUs deployed by top hyperscalers to reach nearly 50% in 2025.
The same Reuters piece cited Google’s use of Arm-based designs in its Axion processors, positioning Arm as a credible efficiency play in modern data centers.
3) Edge AI: Arm is trying to become the default “AI on-device” substrate
In October, Reuters reported that Arm expanded its Flexible Access licensing program to include its Armv9 edge AI platform, aiming to lower barriers for startups and device makers building on-device AI. Arm said more than 300 companies were in the program and had produced 400 manufacturing-ready chip designs.
This is the “a billion little AIs” strategy: not just data centers, but phones, cars, cameras, and industrial devices.
4) Big-name validations: Meta joins the Arm server story
Reuters reported that Meta partnered with Arm to use Arm-based data center platforms for AI ranking and recommendation systems across Facebook and Instagram, and that the companies planned to open-source software improvements to improve compatibility and adoption.
That kind of adoption matters because software ecosystem friction has historically been Arm’s biggest obstacle in server land.
The bear case: what skeptics are worried about
If bulls are selling “Arm everywhere,” bears are selling “Arm is priced for everywhere… but paid like somewhere.”
1) “AI leverage” debate: does Arm capture enough upside?
Goldman’s downgrade is basically an argument about capture. Even if AI spending rises, Arm’s business model may not scale in lockstep the way GPU, memory, or certain infrastructure plays can.
One cited concern: Arm’s royalty exposure to smartphones and the reality that royalty rates and unit growth can limit near-term upside (depending on contract structures and end-market growth).
2) The strategic shift risk: when your customers become your competitors (or vice versa)
A recurring tension: Arm is exploring going beyond licensing into more complete solutions—and potentially even finished chips.
Reuters reported earlier in 2025 that Arm would invest toward developing finished chips, with the CEO declining to detail specific plans at that time.
Reuters also reported Arm hired a former Amazon AI chip director (involved with Trainium and Inferentia) to help push that ambition forward.
Strategically, this could expand Arm’s revenue pool. Socially, it can spook the ecosystem: Arm’s customers don’t love funding a future competitor.
3) Regulatory scrutiny: licensing practices under a microscope
Reuters reported that South Korea’s antitrust regulator investigated Arm’s Seoul offices amid scrutiny of Arm’s licensing practices.
Regulatory attention like that can create headline risk and, in some cases, affect negotiating posture with customers.
4) The open-standard threat: RISC‑V grows up
The most existential long-term question isn’t “Is Arm good?” It’s “Is Arm inevitable?”
Reuters Breakingviews argued that open-standard chips using RISC‑V could become a meaningful force in AI and semiconductors as early as 2026, driven partly by geopolitical neutrality and growing interest from major firms. The piece cited estimates putting RISC‑V at 10.4% penetration (by chip value) in 2024 and suggested the performance gap could narrow significantly by 2027, with the RISC‑V market potentially exceeding $260 billion by 2030.
Arm is still massively entrenched. But investors don’t wait for disruption to finish—they price the possibility of disruption early.
ARM stock forecasts: what analysts are projecting now
Even after the latest wobble, consensus-style aggregates still skew bullish—but with a widening disagreement band.
- MarketBeat’s compilation (as of mid-December) showed an average 12-month ARM price target around $178.30, with targets ranging roughly from $120 to $210.
- A separate Nasdaq.com analyst-note summary referenced an average one-year target around $169.85 (with a wide range of estimates), reflecting how dispersed views are on valuation versus growth.
That wide spread is a signal in itself: Arm is being valued less like a stable “semiconductor utility” and more like an outcome-dependent AI platform.
What to watch next for Arm Holdings stock
Next earnings: early February 2026 is the next big checkpoint
Multiple market calendars currently point to Feb. 4, 2026 as the expected next earnings date window for Arm (noting that some listings describe this as an estimate based on historical schedules).
For investors, the “make-or-break” items tend to be:
- Armv9 and CSS traction (royalty rate uplift)
- data center adoption signals
- licensing momentum and contract mix
- R&D spending trajectory (especially if chip ambitions expand)
AI capex and “bubble” narratives will keep swinging the mood
When the market is anxious about AI spending discipline, high-multiple names can get hit regardless of execution. Reuters’ recent reporting on AI-trade volatility underscores how quickly sentiment rotates when large tech players adjust capex or guidance.
Ecosystem diplomacy: Arm’s toughest engineering problem may be… human
Arm’s best strategic move could also be its riskiest: moving up the stack to capture more value. Every step upward potentially stresses relationships with licensees. That’s not fatal—but it’s never frictionless.
Bottom line: Arm is still a core AI infrastructure story, but the market wants cleaner math
Arm Holdings remains one of the most important companies in modern computing because it sits at the blueprint layer: licensing the architecture that shows up across smartphones, embedded devices, and increasingly, servers.
What’s changing into year-end 2025 is the market’s tone. Investors are not debating whether Arm is relevant. They’re debating how much of the AI boom Arm monetizes, how costly the next strategic phase will be, and whether today’s valuation leaves room for ordinary execution—rather than perfection.
