AAPL, NVDA, TSLA: Analyst Dan Ives Lists the 10 Tech Stocks to Own Now

Wedbush analyst Dan Ives remains constructive on U.S. technology, calling AI in its early innings and forecasting technology capex of $550–$600 billion for 2026 and significant government AI spending. He publishes a 10-stock buy list into year-end — AMD, Alphabet, Apple, CrowdStrike, Meta, Microsoft, Nvidia, Palantir, Palo Alto Networks and Tesla — spanning chipmakers, cybersecurity and mega-cap tech. Wedbush says it is 'firmly bullish' into 2026, and the piece notes Apple carries a consensus Moderate Buy from 35 analysts with an average $289.17 target implying ~3.74% upside.

Analysis

Market structure: The Wedbush thesis implies concentrated winners — datacenter GPU vendors (NVDA, AMD), cloud OS/AI stacks (MSFT, GOOGL) and cybersecurity (CRWD, PANW) — should capture disproportionate revenue from a projected multi-year capex wave into 2026. Expect pricing power for top-tier GPUs and cloud AI services, while legacy on-prem vendors and low-end silicon suppliers face margin pressure; this will likely raise semiconductor lead times and bid up inputs (memory, wafers) for 6–18 months. Cross-asset: a tech-led risk-on skews equities higher, likely pressuring IG spreads +10–30bps and pushing 10y yields +10–30bps if buybacks/capex accelerate; commodity impacts concentrated in copper/energy for data-center buildouts.

Risk assessment: Tail risks include export controls (BIS/US/EU actions within 30–90 days), antitrust enforcement on cloud/ads players, and a demand pullback if enterprise AI ROI proves slower than pilots — any of which could erase 25–50% of forward earnings expectations for mid-cap names. Short-term (days–weeks) volatility around earnings/guidance; medium (3–12 months) depends on product rollouts and government budgets; long-term (12–36 months) dominated by TSMC/Intel capacity and talent costs. Hidden dependencies: hyperscaler procurement cycles, TSMC capacity allocation, and data-center power constraints can bottleneck real adoption.

Trade implications: Favor concentrated, size-controlled longs: NVDA (3% portfolio), MSFT/GOOGL (2% each), CRWD/PANW (1.5% each) funded by trimming cyclicals (industrials/consumer discretionary by 3–5%). Use 3–9 month option structures to control risk: sell 10% OTM cash-secured puts on NVDA for entry or buy 1:1 call spreads to cap premium; calendar spreads on MSFT into next earnings. Pair trade: long CRWD vs short PLTR (net 0.5–1% exposure) to express security software quality dispersion.

Contrarian angles: Consensus underestimates implementation cost and margin erosion from large-scale retraining and inference (driving higher opex for hyperscalers), which could compress cloud margins 200–400bps over 12–24 months. NVDA’s premium pricing and concentration risk is a vulnerability — regulatory or capacity shocks could trigger >30% drawdowns; conversely, AAPL and META are underappreciated optionality plays (device AI and ad monetization) that could outperform if monetization accelerates.

Key Decisions for Investors

• Establish a 3% portfolio long in NVDA via either (A) selling 10% OTM cash-secured puts expirying in 3–6 months (collect premium, max assign risk) or (B) buying a 6-month 1:1 call spread (buy ATM, sell +20% strike) — target 30–50% upside in 6–12 months, stop-loss if position down 20%.
• Add 2% longs in MSFT and GOOGL each (buy shares or 6–9 month calls) to play cloud AI stack; set blended profit target 15–25% over 12 months and trim half position on outperformance >25% to lock gains.
• Construct a cybersecurity pair: go long CRWD 1.5% and PANW 1.5% (equal weight) funded by reducing cyclicals by 3–5%; concurrently short PLTR 1% (shares or buy-to-open puts) to hedge hype — objective: capture 15–25% relative upside in 6–12 months, stop-loss if PLTR outperforms CRWD by >15% in 60 days.
• Rotate portfolio: reduce industrials and consumer discretionary exposure by 3–5% and reallocate to the above tech/cyber longs over 30–45 days to avoid earnings-driven spikes; maintain cash buffer 5% for drawdown buying.
• Monitor regulatory catalysts closely: track US BIS export control updates and EU AI Act milestones over next 30–90 days — if measures explicitly restrict datacenter GPU exports or impose heavy compliance costs, reduce NVDA/AMD exposure by 50% within 5 trading days of rule publication.