Despite Bubble Fears, 93% of AI Investors Say They're Bullish on Long-Term Returns. Here Are 2 Stocks to Hold for a Decade or More.

A Motley Fool survey found 93% of AI-stock investors dismiss concerns of an AI bubble, reflecting strong investor conviction in long-term AI returns. Nvidia reported fiscal 2026 Q3 revenue of $57.0 billion, up 62% year‑over‑year, driven by GPU demand as data centers upgrade for AI workloads and growth opportunities in agentic AI and autonomous vehicles (including a partnership with Uber). Alphabet’s Q3 revenue rose 16% to $102.3 billion, with Google Search revenue up to $56.6 billion from $49.4 billion year‑over‑year, while its AI roadmap includes Waymo expansion in U.S., Tokyo and planned London rollout. Together, the results underscore durable demand for AI infrastructure and services and provide differentiated exposure to the AI-driven secular growth theme.

Analysis

Market structure: Nvidia (NVDA) and Alphabet (GOOGL/GOOG) are primary beneficiaries as AI-driven demand re-routes datacenter capex from CPU-centric to GPU/accelerator-centric architectures; NVDA’s fiscal Q3 revenue +62% to $57B and Google Search revenue $56.6B (Q3) confirm secular demand. Winners include hyperscalers, FPGA/accelerator software stacks and power/UPS vendors; losers are legacy CPU cycle sellers, smaller AI chip startups unable to scale, and low-margin logistic/driver-based marketplaces if autonomous fleets scale.

Risk assessment: Key tail risks are tighter US export controls on advanced AI chips (30–60% downside to NVDA consensus in stressed scenario), antitrust/safety regulation for Google advertising or AI output liabilities, and a demand snapback if hyperscalers saturate capacity. Near-term (days–weeks) sensitivity centers on headline-driven sentiment and options IV; medium-term (3–12 months) on enterprise capex cadence and supply (foundry capacity); long-term (3–5+ years) on competitive custom silicon and regulation.

Trade implications: Tactical long NVDA and GOOGL exposure with volatility-defined options overlays is preferable to naked longs—buy 3–6 month NVDA call spreads 8–15% OTM sized to 1–2% NAV, and sell 1–2 month covered calls on accumulated GOOGL to harvest premium while holding 1–3% core weight. Pair trades: long NVDA vs short a CPU-reliant datacenter ETF or legacy server OEMs to express architecture rotation. Rotate +3–5% from ad-dependent media into semis/cloud infra over next 90 days.

Contrarian angles: Consensus underestimates concentration risk—>50% of NVDA revenue driven by a handful of hyperscalers—so upside is large but binary; the market may be pricing perpetual multiple expansion. Historical parallel: dot‑com era where fundamentals diverged from sentiment—expect episodic 20–40% drawdowns. Unintended consequence: widespread AI helpers could compress CPCs long-term, pressuring ad-based monetization beyond near-term gains.

Key Decisions for Investors

• Establish a 1–2% NAV long position in NVDA now; supplement to 3% NAV if price retraces 10–20% within the next 90 days. Hedge with a 3–6 month call spread (buy 8–12% OTM, sell 20–25% OTM) to cap premium outlay and define downside.
• Accumulate a 1–3% NAV core position in GOOGL over 30 days using dollar-cost averaging. Sell 30–45 day covered calls at ~10% OTM to generate yield while retaining upside; add to position if forward P/E falls below 18 or organic Search growth slows <10% YoY.
• Initiate a pairs trade: long NVDA (1% NAV) vs short 1% NAV of a legacy server/CPU vendor or datacenter ETF heavily weighted to CPU firms to express architecture rotation; target pair profit of 25% or tighten if pair outperforms/worsens by 15%.
• Use options volatility: if NVDA IV spikes >60% (30-day implied), sell 20–40 day call credit spreads 20–30% OTM for income sized to 0.5–1% NAV. Conversely, buy protective puts 3–6 months out if NVDA drops >25% in 30 days.
• Reduce or underweight exposure to driver-dependent mobility names (e.g., UBER) by 1–2% NAV over next 6–12 months; autonomous rollout timelines (Waymo/Uber) and margin impacts are uncertain and could compress legacy marketplace economics.