2 AI Growth Stocks to Buy Now

Amazon and Microsoft are positioning AI as a company-wide growth layer with material impact on their cloud businesses: Amazon reported Q3 sales of $180.2B (+13% YoY) with AWS revenue of $33.0B (+20.2%) and AWS operating income of $11.4B (66% of Amazon operating income), while Amazon's operating cash flow rose to $130.7B but free cash flow fell to $14.8B due to a $50.9B YoY increase in capital expenditures for data centers. Microsoft posted fiscal Q1 revenue of $77.7B (+18% YoY) and operating income of $38.0B (+24%), with Azure and other cloud services revenue up 40% YoY and commercial RPOs up 50% to over $400B; capex was $34.9B in the quarter and is expected to rise further in fiscal 2026. Both companies trade at rich forward P/Es (Amazon ~29x, Microsoft ~30x); the piece flags AI-driven capex as a key execution risk but views the long-term earnings upside as significant, advising caution on position sizing given valuation and spending risk.

Analysis

Market structure: Hyperscalers (MSFT, AMZN) and AI infra suppliers (NVDA, AMD, data‑center REITs with power capacity) are clear winners as enterprise demand for GPU hours and cloud RPOs compresses competition; smaller cloud providers and on‑prem vendors will lose pricing power. Expect sustained tight supply for top‑tier GPUs and rack/power capacity over 6–18 months, supporting higher ASPs for chips and colocation, and structurally higher capex among leaders (MSFT/AMZN capex up to $35B+ quarterly). Cross‑asset: stronger tech cash flows and capex raise rate‑sensitivity — net effect is modest upward pressure on long‑end yields if capex funds via debt, wider implied vols on NVDA/MSFT options, and commodity pressure for copper/power in regional grids hosting clusters.

Risk assessment: Tail risks include regulatory curbs on model training/hosting or export controls on accelerators, a chip supply shock, or a macro enterprise spend pullback; any of these could shave 10–30% off consensus cloud growth over 12 months. Short term (days–weeks) expect earnings‑driven repricings; medium (3–12 months) is execution on capex and RPO conversion; long term (1–3 years) returns hinge on margin recovery as capex converts to revenue. Hidden dependencies: reliance on third‑party silicon (Nvidia) and local grid/power constraints; monitor GPU inventory and regional power curtailments as second‑order limits to growth. Catalysts: major model launches, large multi‑year cloud deals, or concrete regulatory proposals in next 30–90 days.

Trade implications: Favor weighted exposure to MSFT over AMZN due to 40% Azure growth and >$400B RPOs — a 2–4% portfolio overweight to MSFT vs 1–2% tactical exposure to AMZN. Use 9–18 month LEAPS or bull call spreads on MSFT to capture secular upside while selling near‑dated OTM calls on AMZN to fund cost due to higher short‑term FCF pressure. Add selective long exposure to NVDA/semis (1–2%) for supply tightness but hedge with 3–6 month protective puts given regulatory/volatility risk. Rotate capital toward utilities/energy infra names servicing hyperscalers if data shows power shortfalls in next 6–12 months.

Contrarian angles: The market understates the margin drag from aggressive capex — FCF for AMZN fell to ~$14.8B; if capex stays +$30–50B YoY, near‑term EPS could lag expectations and create 10–20% downside. Conversely, consensus may underpay for MSFT’s RPO conversion — if commercial RPO converts at even 60% of bookings within 12–18 months, revenue upside could surprise by mid‑teens percent. Historical parallel: cloud cycles have seen front‑loaded capex then multi‑year margin expansion (2013–2018); if this repeats, earlier entrants (MSFT) capture disproportionate share but only after a 6–24 month execution window.

Key Decisions for Investors

• Establish a 2–4% portfolio long position in MSFT over 2–6 weeks; prefer 9–18 month LEAPS (or 1:1 bull call spreads) to capture Azure/RPO conversion. Trim or take 50% profits if MSFT rallies >25% within 6 months or forward P/E >35.
• Add a tactical 1–2% long in AMZN but use a buy‑write or collar to protect downside (sell 6–12 week OTM calls to fund cost). Only add incremental exposure if FCF recovers above $25B trailing 12 months or AWS growth accelerates above 25% YoY.
• Execute a relative trade: long MSFT (2%) / short AMZN (1%) to express Azure outperformance; rebalance quarterly and unwind if Azure growth decelerates to <25% YoY or AWS growth exceeds 30% YoY.
• Allocate 1–2% to NVDA/semiconductor exposure for supply tightness but buy 3–6 month ATM puts (or use collars) to hedge regulatory or demand shock risk; reduce this sleeve if GPU supply ramps >30% in next 6 months.
• Monitor regulatory activity (US/EU model‑hosting rules, export controls) and quarterly GPU inventory reports over the next 30–90 days; if substantive restrictive proposals surface, reduce net tech exposure by 5–10% within 30 days.