Google Plans to Double Capex Spending as Cloud Growth Soars 48%

Alphabet reported Q4 revenue of $113.8 billion, up 18% year-over-year (17% in constant currency), and diluted EPS of $2.82, a 31% increase, beating consensus revenue of $111.48 billion and EPS of $2.64. Google ad revenue rose 14% (search +17%, YouTube ~9%), while Google Cloud surged 48% to $17.7 billion with operating margins improving to 30.1% from 23.7% in Q3; management sold more than 8 million paid seats of Gemini Enterprise and reports 750 million monthly active users for the Gemini app. Management signaled heavy investment to meet AI/cloud demand, guiding to roughly doubled capex of $175–$180 billion for 2026 (vs. ~$91 billion last year), and the shares have rallied ~65% over the past year while trading under 30x forward earnings.

Analysis

Market structure: Alphabet (GOOGL/GOOG) is a clear winner — 48% cloud growth to $17.7B, rising Google Cloud margins (30.1%) and Gemini traction (8M paid seats, 750M MAU) signal growing pricing power for AI compute and enterprise software; immediate beneficiaries include NVDA and data‑center vendors, while smaller cloud providers and legacy ad‑dependent media (YouTube underperformance) face margin pressure. Competitive dynamics: Google is closing the gap with MSFT (Azure growth 39% vs Google 48%), which should reshape share dynamics in enterprise AI over 12–36 months and allow Alphabet to command higher ARPU for AI services as supply remains constrained.

Risk assessment: Tail risks include aggressive antitrust/AI regulation (legal action or forced structural changes), GPU/TPU supply shocks, and execution risk on a planned capex jump to $175–180B in 2026 (vs ~$91B prior) that could depress FCF conversion in 2025–2026. Time horizons matter: expect immediate positive sentiment (days–weeks), operational FCF drag in 6–18 months as capex hits, and potential margin leverage after 18–36 months if AI demand sustains; hidden dependencies include power/permitting and custom‑chip availability.

Trade implications: Favor selective longs in GOOGL and NVDA and protect with defined‑risk options: establish core long GOOGL exposure with LEAP call spreads to capture 25–40% upside over 12–24 months while capping cost, and buy NVDA (supply beneficiary) for 6–12 month momentum. Consider a relative pair (long GOOGL, short MSFT) sized neutral to hedge macro beta for 6–12 months and trim high‑beta ad‑dependent small caps if ad growth falls below 5% QoQ.

Contrarian angles: The market underestimates capex execution and FCF risk — doubling capex could produce overcapacity if AI demand softens, making multi‑year ROIC the real arbiter. Regulatory risk is underpriced: a material antitrust outcome or AI safety rules in 12–24 months could compress multiples >20%; conversely, successful product monetization (Gemini enterprise scale) would justify a move to 30–35x forward earnings, so watch adoption metrics and enterprise contracts closely.

Key Decisions for Investors

• Establish a 2–3% portfolio long position in Alphabet (GOOGL) within the next 5 trading days; add up to +1% on a pullback of >5% from current levels. Use position size limits to keep portfolio beta in check given capex risk.
• Buy a defined‑risk bullish LEAP on GOOGL (12–24 month call vertical) sized to express conviction equal to 1–2% notional—target a 25–40% upside; alternatively sell OTM puts (15–20% OTM) to acquire stock if willing to own at a discount, only if implied vol premium < historical 90‑day realized vol + 20%.
• Establish a 1–2% long position in NVDA over the next 30 days to capture ongoing AI compute demand; use 6–12 month time horizon and consider 1:1 covered call overlays if NVDA rallies >20% to lock gains.
• Implement a market‑neutral pair trade: long GOOGL 0.75% / short MSFT 0.75% for 6–12 months to play relative cloud/AI share gains; close if Google cloud growth reverts <30% YoY or MSFT reports >50% cloud acceleration.
• Reduce exposure by 2–4% in ad‑dependent small caps and consumer media names if aggregate ad revenue growth for the sector falls below +5% QoQ in the next two earnings cycles; redeploy proceeds into AI infrastructure names (GOOGL, NVDA, data‑center suppliers).