Arm vs NVIDIA: Analysis of Performance Divergence in the AI Chip Era

Based on the obtained data, I can now conduct a comprehensive analysis of the reasons for the divergent stock performance between Arm and NVIDIA.

Indicator	Arm (ARM)	NVIDIA (NVDA)
Current Price	$107.16	$184.57
Daily Change	-3.58%	-0.20%
Market Cap	$113.2B	$4.49T
Price-to-Earnings (P/E) Ratio	137.38x	45.80x
52-Week Performance	-21.82%	+38.53%
1-Month Performance	-18.13%	+5.45%

[0]

Today (January 13), Bank of America downgraded Arm’s rating from “Buy” to “Neutral”[1][2]. This marks another major bearish signal following Goldman Sachs’ downgrade of Arm from “Neutral” to “Sell” in December 2025[0].

Arm’s latest quarterly earnings report (Q2 FY2026) shows[0]:

• Actual EPS
  : $0.15 vs. expected $0.36 (
  a staggering -58% negative surprise
  )
• While revenue slightly beat expectations ($1.14B vs. expected $1.12B), profitability fell far short of expectations

Arm currently has a sky-high P/E ratio of

, compared to NVIDIA’s [0]. In a high-interest rate environment, high-valuation stocks face greater pressure from valuation compression.

---

Competitive Advantage	Specific Performance
Market Dominance	Data center business accounts for 87.9% of total revenue , with Q3 FY2026 revenue reaching $57B[0]
CUDA Ecosystem Moat	Integrated hardware and software ecosystem with extremely high switching costs
Profitability	Net profit margin 53% , operating profit margin 59% [0]
Product Iteration	Continuous advancement of the Blackwell architecture, with in-depth collaboration with Google Cloud[3]
Geopolitics	China has restrictions on H200 chip purchases, but approvals are still possible under special circumstances[4]

Competitive Advantage	Specific Performance
Architecture Licensing Model	Collects chip design licensing fees and royalties, operates a light-asset business model
Mobile Market Leader	Virtually all smartphone chips are based on the Arm architecture
Edge AI Strategy	Launched a new Physical AI division to expand into robotics and edge computing[5]
Energy Efficiency Advantage	Edge AI devices prioritize low power consumption, a natural strength of the Arm architecture

---

Dimension	NVIDIA	Arm
Revenue Model	Direct sales of high-value GPU chips	Licensing fees + royalties (indirect benefit)
AI Revenue Conversion	Immediate, direct, and large-scale	Slow, indirect, and dependent on terminal shipments
Price-to-Earnings Ratio	46x (high but reasonable)	137x (excessive overestimation of growth expectations)
Revenue Scale	Quarterly revenue of $57B	Quarterly revenue of $1.14B
Net Profit Margin	53%	19%
Upside Potential from Analyst Target Prices	+45.5%	+68% (but multiple downgrades in recent times)

[0]

: Arm’s licensing model means that growth in AI chip demand must go through multiple stages to translate into its revenue—chip manufacturers need to adopt the Arm architecture to design AI chips → chip mass production → terminal device sales → Arm collects royalties. In contrast, NVIDIA sells GPUs directly to cloud providers and enterprises, so its revenue growth is almost synchronized with AI demand.

---

1. Consistent Earnings Beats
   : Q3 FY2026 EPS beat expectations by 3.17%, while revenue beat expectations by 3.81%[0]
2. High Analyst Recognition
   : 75.9% of analysts give a Buy or Strong Buy rating[0]
3. Relatively Reasonable Valuation
   : A 46x PE ratio is acceptable for a company with annual revenue growth of over 40%
4. Uncertainties in the Chinese Market
   : News of approved H200 purchases in some cases provides a positive catalyst[4]

---

Arm’s pre-market/intraday plunge is mainly due to:

• The direct impact of
  Bank of America’s rating downgrade
• A crisis of confidence triggered by
  last quarter’s EPS falling far short of expectations
• The
  137x P/E ratio
  being unsustainable in the current market environment

NVIDIA’s more stable performance is due to:

• Being the
  direct market leader
  in AI training/inference chips
• Consistent earnings beats
  that validate its growth logic
• A
  relatively reasonable valuation
  that provides a better margin of safety

Against the backdrop of growing demand for AI chips,

—there are fundamental differences in their revenue transmission efficiency and valuation logic.

---

[0] Jinling API Data
[1] Yahoo Finance - “AMD upgraded, Arm downgraded: Wall Street’s top analyst calls” (https://finance.yahoo.com/news/amd-upgraded-arm-downgraded-wall-144008117.html)
[2] Seeking Alpha - “Arm’s rating cut at BofA ahead of earnings results” (https://seekingalpha.com/news/4538746-arms-rating-cut-at-bofa-ahead-of-earnings-results)
[3] YouTube - “Alphabet and NVIDIA Bring Agentic and Physical AI to Global Industries” (https://www.youtube.com/watch?v=UXyv0QvbsJY)
[4] Seeking Alpha - “China approves tech companies buying Nvidia H200 GPUs in some cases” (https://seekingalpha.com/news/4538859-china-approves-tech-companies-buying-nvidia-h200-gpus-in-some-cases-report)
[5] Yahoo Finance - “Arm Holdings (ARM) Expands in the Robotics Industry With Physical AI Unit” (https://finance.yahoo.com/news/arm-holdings-arm-expands-robotics-174728081.html)