U.S. AI Tech Stocks — Bubble or Opportunity?
Examining the valuations of AI beneficiaries like NVIDIA and Microsoft, and presenting long-term investment criteria.
The AI Rally, Now in Its Second Year
The AI tech rally sparked by the ChatGPT frenzy in early 2024 continues into 2026. The excess return of the Nasdaq 100 over the S&P 500 is approaching all-time highs. The question is: can this momentum sustain?
Key Stock Valuation Check
NVIDIA (NVDA)
- P/E: ~45x (TTM basis)
- Data center revenue share: 87% of total
- Blackwell GPU backlog: 6–12 months
Views are split between those who see it as overvalued and those who argue AI infrastructure investment is in a structural growth phase that justifies the premium.
Microsoft (MSFT)
- P/E: ~35x
- Azure AI service revenue growth: YoY +42%
- Copilot enterprise subscriptions growing rapidly
Core Metrics in the Bubble Debate
During the 2000 dot-com bubble, the Nasdaq's peak P/E was roughly 175x. The current level (~35x) is high but backed by fundamentally different earnings.
Differences:
- Earnings-backed: Both NVIDIA and Microsoft are delivering real revenue and profit growth
- Strong cash flows: Major Big Tech companies maintain FCF margins above 20%
Similarities:
- Concentration risk: The Magnificent 7 account for 33% of S&P 500 market cap
- Rate sensitivity: Rising long-term bond yields increase discount rates for high-P/E stocks
Long-Term Investment Strategy
- Dollar-Cost Averaging: Enter positions in tranches during volatile periods
- Non-AI Diversification: Balance the portfolio with energy, healthcare, and other sectors
- Profit-Taking Rules: Set rules to take partial profits when target P/E is exceeded