Are Amazon and Alphabet Among the Best Stocks to Buy Now?

AWS was Amazon's fastest-growing segment, rising 24% YoY in Q4 and accounting for 50% of Amazon's operating profit in Q4 (66% in Q3), its best quarter in over three years, as AI demand boosts cloud growth. Google Cloud grew 48% YoY in Q4 with a roughly 30% operating margin. Both Alphabet and Amazon trade at a premium to the S&P 500 forward P/E (~21.7x) and are off recent highs (~10% for Alphabet, ~16% for Amazon); the author acknowledges strong cloud-driven fundamentals but prefers Microsoft and Nvidia as better values.

Analysis

Scale in hyperscaler AI spending is creating a bifurcation: capital-light software capture vs capital-heavy infrastructure capture. Hardware and systems vendors that sit upstream of model training and inference (accelerators, high-bandwidth networking, power and cooling OEMs) will likely see revenue growth that is both stickier and lumpier versus cloud platform revenue, which can be lumpy but is easier to monetize incrementally through software services and ads. Expect margin dispersion to widen across the stack — suppliers with fixed-cost factories will show volatile profits tied to multi-quarter OEM order book shifts, while software-layer incumbents will enjoy smoother operating leverage.

Key tactical risks: model efficiency improvements or a shift to on-prem inference could materially shorten the current procurement cycle and deflate server-level ASPs within 6–18 months; conversely, a sustained wave of large-scale model deployments will force a fresh capex cycle among hyperscalers, favoring GPU/ASIC suppliers and colo providers. Regulatory or corporate governance actions that limit data access or force different revenue-sharing with partners would compress platform multiples faster than infrastructure multiples, creating asymmetry in short-term returns. Watch guidance cadence and procurement lead-times from hyperscalers as a leading indicator — order books change months before revenue shows up.

The consensus is underweighting concentration risk: compute demand may re-concentrate around a smaller set of hyperscalers, amplifying supplier bargaining power but also creating single-point-of-failure macro exposure. That dynamic makes pure-play hardware suppliers a levered but directional way to express the AI cycle, while owning diversified software/platform exposure hedges idiosyncratic capex shocks. Time your entries around hyperscaler earnings and capex guidance windows where visibility shifts from noisy to directional over a 3–12 month horizon.

Key Decisions for Investors

• Directional hardware spread: Buy a 9–12 month call spread on NVDA sized to 1–1.5% of NAV (long-call nearer-term, sell higher-strike call) — objective: 60–120% return if AI-driven GPU demand continues; max loss = premium paid, take-profits at ~2x, reassess on next quarterly capex guides.
• Cloud margin pair (12-month horizon): Long MSFT equity 3–5% NAV, financed by a 2–3% NAV short position in AMZN (cash-secured or put spread) — target 30–40% relative outperformance if Microsoft sustains enterprise pricing power; hard stop-loss on the pair at 15% adverse move to limit platform-concentration risk.
• Optional asymmetric bet: Small allocation (0.5–1% NAV) to long-dated INTC LEAPS (12–18 months) as optionality on a process/architecture catch-up — upside >3x if Intel reclaims meaningful AI silicon share, capped loss = premium; trim into any sustained 2–quarter order-book recovery signal.
• Event risk hedge: Ahead of hyperscaler earnings, buy short-dated protective puts on NVDA or MSFT equal to ~0.5% NAV to protect against a sudden compute-demand re-rating (use 4–8 week expiries); cost accepted as insurance against a rapid sentiment reversal tied to capex disappointment.