Morgan Stanley's Optimistic Assessment of AI Demand at CES 2026 and Investment Strategy Analysis

Morgan Stanley’s research report released ahead of CES 2026 expresses a

stance on the prospects of AI demand, based on the following core observations:

• Hyperscale enterprises (Microsoft, Meta, Alphabet, etc.) are expected to invest over $500 billion in building AI infrastructure in 2026
  , including data centers, networks, and computing power facilities [1]
• Data center M&A transactions reached approximately $70 billion in 2025, with half allocated to short-term assets (GPUs, CPUs) and half to long-term assets (land, power, cooling systems) [6]
• Alphabet has raised its 2025 capital expenditure forecast to $85 billion and expects it to climb further in 2026 [6]

Morgan Stanley strategist Andrew Sheets specifically noted that

:

• Chip costs and power costs continue to rise
• The huge demand for energy and advanced chips from large tech companies is pushing up project costs
• Memory chip cost inflation will drive up AI product prices, potentially reducing investor returns and affecting capital inflows [1]

According to the analysis by the Consumer Technology Association (CTA), CES 2026 demonstrated the following

[2][3]:

Trend	Core Content	Investment Opportunities
Intelligent Transformation	AI is no longer just an overlay, but infrastructure embedded in manufacturing, logistics, healthcare, and urban systems	Industrial automation, AI-optimized systems, intelligent logistics
Physical AI	Robots, autonomous vehicles, and industrial machines move from prototypes to actual operations	Robotics technology, autonomous driving, sensors
AI Everywhere	AI is embedded in all industries from smart manufacturing to personalized healthcare	Vertical industry AI applications, AI+SaaS

, emphasizing technology, partnerships, and employee training for AI-ready production facilities [3]. This marks a key turning point in AI investment penetration from pure technology fields to traditional industries.

The chart shows the investment priority of AI sub-sectors, value capture across investment cycles, valuation comparison of leading stocks, and portfolio allocation recommendations [0]

Based on CES 2026 trends and Morgan Stanley’s analysis, I have constructed the following

:

: This is the “water seller” of AI demand; regardless of how downstream applications evolve, infrastructure investment is indispensable.

• AI Chips & Foundry
  • TSM (Taiwan Semiconductor Manufacturing Company)
    : Morgan Stanley analyst Charlie Chan expects that if TSMC guides for 25% YoY revenue growth in 2026 and increases the CAGR of AI semiconductor foundry revenue (2024-2029) to 60%, its stock price may approach the target price of NT$1,888 (current $330.14) [0]. A more optimistic scenario is an AI semiconductor CAGR of 75% and revenue growth of 30% [0]
  • NVDA (NVIDIA)
    : Current market capitalization of $4.66 trillion, data center business accounts for 87.9% [0], BofA calls it “best positioned” as it optimizes the entire AI stack from networks to GPU workload management [7]
  • AMD
    : Data center business accounts for 47.0% [0], target price of $300 (37.5% upside potential) [0]

• High Bandwidth Memory (HBM)
  • MU (Micron)
    : Stock price rose 5.53% today to $329.40, up 72.50% in 3 months and 231.86% in 1 year [0], benefiting from the urgent demand for HBM in AI data centers
  • BofA is bullish on HBM, advanced logic chips, and supporting infrastructure as main growth engines [4]

• Short-term (2026Q1-Q2)
  : Focus on AI chips and foundry, value capture potential of 90% [0]
• Mid-term (2026Q3-Q4)
  : Value capture potential drops to 85%, need to pay attention to valuation pressure [0]

: Data centers are becoming the “new oil fields” of AI capital; Big Tech is building a “second economy” around land, power, and cooling systems [6].

• Hyperscale Cloud Service Providers
  :
  • MSFT (Microsoft)
    : Market capitalization of $3.52 trillion, P/E ratio of 33.70x [0], OpenAI partner, building its own data centers and developing its own chips [5]

• Data Center Infrastructure
  :
  • Power, cooling, network equipment suppliers (e.g., Amphenol, APH) [4]
  • Data center REITs and related utility stocks benefit from AI data center demand [6]

: BofA analyst Vivek Arya pointed out that CES 2026 will strengthen the new driving force of [7], which represents an important shift of AI from the cloud to the edge.

• Edge AI Chips
  : Companies like Intel (Panther Lake) are trying to integrate real AI hardware at every layer of the device stack [7]
• AI-Driven Consumer Electronics
  : AI processors in smartphones, PCs, wearable devices
• Network Infrastructure
  : 5G/6G, edge computing nodes

• Mid-term (2026Q3-Q4)
  : Value capture potential rises to 75% [0]
• Long-term (2027+)
  : Continues to benefit, value capture potential of 75% [0]

: As AI infrastructure improves, the application layer begins to deliver value, but faces commoditization risks.

• Enterprise AI Applications
  : AI-driven productivity tools, customer service automation
• Vertical Industry AI
  : Healthcare AI, financial AI, legal AI
• AI Agents
  : Intelligent agents that can autonomously plan and execute tasks [5]

• Mid-term (2026Q3-Q4)
  : Value capture potential of75% [0]
• Long-term (2027+)
  : Value capture potential of90%, becoming the main growth engine [0]

###3.5 Fifth Tier: Physical AI (Investment Priority:75 Points)

: At CES2026, autonomous driving and robotics technology took center stage; investors bet that AI will revitalize an industry plagued by slow progress, high costs, safety accidents, and regulatory scrutiny [2].

• Autonomous Driving
  : L4/L5 technologies from Tesla, Waymo, etc.
• Industrial Robots
  : AI-driven smart manufacturing and logistics robots
• Humanoid Robots
  : Tesla Optimus, Figure AI, etc.

• Mid-term (2026Q3-Q4)
  : Value capture potential of65% [0]
• Long-term (2027+)
  : Value capture potential of85%, needs more time to mature [0]

###4.1 Core Principles

According to BofA’s analysis,

; previous platform cycles lasted 5-10 years, and valuations are still attractive (AI leaders have a PEG of about1x vs SPX/large-cap stocks of1.5x-2.0x) [7]. Investors should focus on:

1. Utilization Level
   : Actual utilization rate of data centers and GPUs
2. Cloud Free Cash Flow Buffer
   : Financial health of Big Tech
3. AI Adoption Trends
   : Rather than “bubble noise, LLM commoditization, reduced training costs, Chinese competition” [7]

###4.2 Portfolio Allocation Recommendations

[0]:

• NVDA
  (25%): Data center accounts for 87.9% [0], analyst target price of $257.50 (34.5% upside potential) [0], 73.4% of analysts give “Buy” rating [0]
• TSM
  (20%): AI semiconductor foundry leader, Morgan Stanley is bullish on 25-30% growth in2026 [0]
• MSFT
  (15%): OpenAI ecosystem, AI application and cloud service leader

: Large market capitalization, good liquidity, high certainty of benefiting from AI infrastructure investment

• MU
  (10%): Explosive demand for HBM, up72.50% in3 months [0], 80.6% of analysts give “Buy” rating [0]
• AMD
  (10%): Data center business accounts for 47% [0], target price of $300 (37.5% upside potential) [0]
• AVGO
  (5%): Network infrastructure and AI accelerators

: High growth potential, benefits from specific AI sub-sectors (HBM, AI accelerators)

• Robotics/Autonomous Driving
  (8%): TSLA, GOOGL Waymo
• AI Software & Applications
  (5%): CRM, SNOW and other AI+SaaS
• Small-Cap High-Growth Stocks
  (2%): EVLV, SOUN, etc. [4]

: High volatility, high return potential, requires stricter profit-taking and stop-loss

###4.3 Risk Management & Hedging

Based on Morgan Stanley’s inflation warning, investors need to manage the following risks:

Risk Type	Hedging Strategy
AI-Driven Inflation	Allocate some defensive assets (utilities, consumer staples) [1]
Valuation Correction	Build positions in batches, avoid chasing highs, focus on targets with PEG ratio <1.5x [7]
Geopolitical Risk	Diversify supply chain risks, focus on targets outside the US and Taiwan
Technology Route Failure	Do not over-concentrate bets on a single technology route

###5.1 Core Insights from Morgan Stanley’s Assessment

Morgan Stanley’s optimistic assessment of AI demand at CES2026 reveals the

:

1. Demand Side
   : AI demand is “soaring”, as evidenced by the $500 billion investment plan of hyperscale enterprises [1]
2. Cost Side
   : Chip and power cost inflation has become an unavoidable resistance, which may squeeze profit margins [1]

This contradiction actually

: regardless of how downstream applications evolve, demand for chips, foundries, and data centers will continue.

###5.2 Next-Stage Investment Themes (By Priority)

:

1. AI Infrastructure
   (NVDA, TSM, MU): Directly benefits from the implementation of the $500 billion investment plan
2. Data Center REITs and Utilities
   : Rising demand for power and cooling [6]

:

1. Edge AI & Device-Side AI
   : Technologies demonstrated at CES2026 begin to commercialize [7]
2. AI Application Software
   : As infrastructure improves, the application layer begins to deliver value

:

1. Physical AI
   (Robotics, Autonomous Driving): Moves from prototypes to large-scale deployment
2. AI Agents
   : Redefines productivity tools [5]

###5.3 Actionable Investment Recommendations

Investor Type	Recommended Strategy	Key Targets
Conservative	Focus on core infrastructure, 70% allocation to NVDA, TSM, MSFT	NVDA, TSM, MSFT
Balanced	60% core +25% growth +15% thematic	NVDA, TSM, MU, AMD
Aggressive	40% core +40% growth +20% thematic (including small-caps)	MU, AMD, EVLV, SOUN

(from BofA recommendations) [7]:

• Data center and GPU utilization rates
• Cloud free cash flow
• AI adoption rate (rather than publicity)
• Impact of inflation on cost structure

---

[0] Gilin API Data - Company Overview, Real-Time Quotes, Technical Analysis of NVDA, TSM, MSFT, AMD, MU and AI Investment Strategy Framework Charts
[1] Reuters - “AI-driven inflation is 2026’s most overlooked risk, investors say” (https://www.reuters.com/world/asia-pacific/ai-driven-inflation-is-2026s-most-overlooked-risk-investors-say-2026-01-05/)
[2] The Drum - “CES 2026: The 3 megatrends set to define this year’s show, according to CTA” (https://www.thedrum.com/news/ces-2026-the-three-megatrends-set-to-define-this-year-s-show-according-to-cta)
[3] China Daily - “AI-driven innovation to transform manufacturing, health and mobility” (https://www.chinadaily.com.cn/a/202601/05/WS695b8686a310d6866eb320fc.html)
[4] Yahoo Finance - “3 Semiconductor Stocks Well-Poised for a Comeback in 2026” (https://finance.yahoo.com/news/3-semiconductor-stocks-well-poised-124000935.html)
[5] Forbes - “10 AI Predictions For 2026” (https://www.forbes.com/sites/robtoews/2025/12/22/10-ai-predictions-for-2026/)
[6] Yahoo Finance - “AI spending is surging, but a hidden risk is getting overlooked” (https://finance.yahoo.com/news/ai-spending-surging-hidden-risk-013700158.html)
[7] Yahoo Finance - “BofA says these are the best 7 AI chip stocks to own in 2026” (https://finance.yahoo.com/news/bofa-says-best-7-ai-161543499.html)