3 Stock-Split Stocks to Buy and Hold for at Least a Decade

Amazon, Netflix and Nvidia are highlighted as long‑term buys on strong AI and advertising tailwinds, robust margins and recent stock splits. Amazon reported Q3 net sales of $180.2 billion, operating income of $17.4 billion, AWS revenue at ~$33 billion (up ~20%) and advertising revenue of $17.7 billion (up 22%), while Prime membership exceeds 240 million. Netflix posted Q3 2025 revenue of $11.5 billion (up 17%), a 28% operating margin and $2.7 billion in free cash flow with a projected ~$9 billion FCF for 2025, and Nvidia delivered record Q3 2026 revenue of $57 billion (up 62%) with data center sales of $51.2 billion (up 66%) and an order backlog cited at $500 billion through 2026, underscoring AI‑driven demand for GPUs and software ecosystem lock‑in.

Analysis

Market structure: Winners are NVDA (data‑center GPUs, ~80–90% share), AWS/AMZN (infrastructure + custom chips), and NFLX (ad tier + high FCF) as AI and ad monetization concentrate spend into a few platforms. Losers include mid‑tier ad networks, legacy media (WBD unless acquired), and non‑CUDA GPU vendors as pricing power and network effects centralize. Tight GPU order backlog (~$500bn) signals demand >> supply for 12–24 months, keeping hardware prices and lead times elevated; risk‑on flows should compress corporate credit spreads and lift tech equities while adding pressure to electricity/commodity inputs for data centers.

Risk assessment: Tail risks include US/China export controls on advanced GPUs, aggressive antitrust scrutiny of AWS/AMZN ads, and a surprise content/M&A write‑down at NFLX; each could knock 20–40% off implied valuations if realized. Near term (days–weeks) reactions will hinge on earnings/announcements and any export‑control headlines; medium term (3–12 months) on supply ramp (TSMC/Nvidia) and ad revenue cadence; long term (years) on sustained AI workload growth and AWS chip adoption. Hidden dependencies: heavy reliance on CUDA lock‑in, hyperscaler ordering cycles, and power/capex constraints that can compress margins.

Trade implications: Direct plays—establish measured longs: NVDA 2–4% portfolio (buy on pullback ≥8–12%), AMZN 3% overweight for AWS+ads (use 6–9m call spreads), NFLX 1–2% tactical long into ad growth (put a 25% stop). Pair trades—long AMZN vs short META (FB) to express ad‑share shift; long NVDA vs short AMD/INTC to capture secular share gains. Options—buy NVDA 3–6m call spreads or buy 6m LEAP calls with 40–60% roll‑down protection; hedge with 3m puts if NVDA rallies >50% from entry.

Contrarian angles: Consensus underestimates margin cyclicality—if AWS monetizes Trainium aggressively, NVIDIA ASPs could compress; conversely, a swift supply increase (TSMC capacity expansion) could unwind the $500bn backlog faster than expected. Valuations for NVDA look rich; consider hedging large long positions with 20–30% notional protective puts. Historical parallels: semiconductor runs driven by architecture leadership often retrace when competitors or regulation alter economics; avoid full concentration risk despite strong fundamentals.

Key Decisions for Investors

• Establish a 2–4% portfolio long position in NVDA on a disciplined basis: accumulate on pullbacks of 8–12% from current levels; use 3–6 month bull call spreads (e.g., buy 1 ATM call, sell 1 +30% strike) to cap cost and target ~40–80% return.
• Add a 3% overweight position in AMZN to capture AWS AI tailwinds and ad growth; implement 6–9 month vertical call spreads (limit cost to <1.5% portfolio exposure) and set a trim rule: reduce exposure by 50% if AWS YoY growth decelerates below 15% in two consecutive quarters.
• Initiate a relative‑value pair: long AMZN (3%) vs short META (2%) to express ad dollar share migration; size so net delta ≈ 0 and rebalance monthly; take profits on a 25–35% differential move in favor of AMZN.
• Keep a tactical 1–2% long NFLX position pre‑earnings if ad revenue growth >40% YoY; hedge downside with 3‑month puts (cost budgeted ≤0.5% portfolio). If NFLX announces a large acquisition (e.g., WBD) with >$10bn purchase price, sell 50% of the position within 2 trading days to lock gains and reassess integration risk.
• Set hard risk triggers to protect concentrated tech exposure: if US export controls broaden to include datacenter GPUs or if NVDA guidance misses by >10% on revenue, cut NVDA exposure by 50% within 48 hours; allocate freed capital to diversified tech ETFs or short‑dated corporate bond hedges.