Back to News
Market Impact: 0.25

3 Artificial Intelligence Stocks You Can Buy and Hold for the Next Decade

Artificial IntelligenceTechnology & InnovationCompany FundamentalsCorporate EarningsAnalyst EstimatesInsider TransactionsM&A & RestructuringTrade Policy & Supply Chain
3 Artificial Intelligence Stocks You Can Buy and Hold for the Next Decade

Amid geopolitical uncertainty and trade tensions driving volatility in equities, the piece highlights Meta, Broadcom and Palantir as buy-and-hold AI exposures based on fundamentals and secular AI demand. Meta commands a 3.3 billion daily-active-user footprint and 700 million monthly users for Meta AI, is planning $60–$65 billion of 2025 capex, trades at 22.3x earnings (5‑yr avg 25.3x) with RBC forecasting 15–20% long-term EPS growth. Broadcom reported AI revenue up 77% YoY to $4.1 billion in Q1 FY25, cites a $60–$90 billion SAM from three hyperscalers by 2027, trades at 27.8x forward (5‑yr avg 70.7x) with analysts modeling ~26.6% long-term earnings growth. Palantir, differentiated by its ontology and new AIP platform, is ~30% below its Feb 2025 high, has seen CEO insider sales (~$1.9bn), trades at ~161.3x forward while analysts forecast ~21.4% revenue and 27.7% earnings CAGR.

Analysis

MARKET STRUCTURE: Winners are AI infrastructure and data-center plays (AVGO, selected semiconductor suppliers, hyperscalers) and ad-monetization leaders with entrenched audiences (META); losers are high-multiple, concentration-risk AI software names (PLTR) and smaller model-only startups that lack scale. Pricing power shifts toward suppliers of XPUs, networking and virtualization stacks (Broadcom/VMware) as hyperscalers internalize scale advantages; advertising yield recovery at Meta reduces CAC pressure for many consumer apps. Supply-demand: persistent GPU/XPU demand and multi-year data‑center buildouts support pricing of accelerators and copper/energy inputs for 2025–2027; near-term supply tightness should favor incumbents. Cross-asset: stronger tech capex expectations lift IG spreads modestly (worse for cyclical sovereigns if risk-off); implied vol will remain elevated for high-beta AI names (PLTR IV > index) while USD moves hinge on tariff/regulatory headlines affecting flow to tech EM suppliers.

RISK ASSESSMENT: Tail risks include rapid AI regulation (model/data restrictions) hitting monetization, a sharp cut in U.S. defense spend shaving PLTR revenue by >20% within 12 months, or capex overruns at Meta that pressure FCF if 2025 spend exceeds $65bn. Time windows: days—earnings, insider sales and tariff headlines; weeks–months—contract renewals and hyperscaler design wins; 12–36 months—realization of SAM targets (Broadcom’s $60–90bn by 2027). Hidden dependency: Broadcom’s SAM concentration (3 hyperscalers driving majority of near-term upside) creates single-client revenue risk; Meta’s ad model depends on retention of 3.3bn DAUs amid regulatory privacy swings. Catalysts: model-adoption cycles, major hyperscaler procurement decisions, and quarterly guidance revisions.

More News