Billionaire Philippe Laffont Is Selling Nvidia and Buying This Other Magnificent Chip Stock Instead

Coatue Management reduced its Nvidia stake by 14% in Q3 (selling ~1.6 million shares) while increasing its Alphabet position by 259% (adding ~5.2 million shares), leaving Nvidia as the fund's eighth-largest holding alongside other AI leaders. The move reflects a rotation from pure-play semiconductors and data-center exposure into a vertically integrated AI play (Alphabet’s Gemini, Google Cloud deals with OpenAI/Meta, Anthropic partnership using TPUs, and Waymo), with the piece noting Alphabet's recent valuation expansion even as it remains relatively cheap on forward P/E versus peers. The trade is framed as prudent rebalancing amid the AI infrastructure buildout and broader investor interest (including an initial Buffett position), suggesting potential follow-through but also caution given stretched valuations.

Analysis

Market structure: The 13F-driven rotation from NVDA into GOOGL signals marginal re-allocation from pure-play accelerators to vertically integrated AI platforms (search, cloud, TPUs, apps). Direct beneficiaries: GOOGL, GCP partners (Anthropic exposure), and integrated software/cloud names (MSFT, AMZN); marginally negative flow for discrete GPU/accelerator suppliers (short-term pressure on NVDA relative returns). Cross-asset effects: risk-on into megacaps should compress equity vol and push real yields modestly higher; energy/equipment suppliers to hyperscalers (power, advanced packaging) see steadier demand outlook.

Risk assessment: Primary tails are regulatory action against Google (antitrust/data rules) or renewed chip export controls limiting advanced node access, each capable of 20–40% downside in impacted names. Timeline: immediate (days) — 13F momentum; short-term (weeks–months) — earnings, product releases (Gemini, Nvidia Qs); long-term (years) — structural share shifts if TPUs or custom silicon materially displace GPUs. Hidden dependency: GCP’s AI credibility hinges on multi-year enterprise contracts (not one-off Anthropic/OpenAI deals) and on TPU benchmark parity vs. Nvidia for a wider workload set.

Trade implications: Tactical play is to capture rotation while hedging technological moat risk — favor modest GOOGL convexity and revenue-exposure to cloud versus trimming but not exiting NVDA. Use relative-value (long GOOGL, hedge with short-dated NVDA calls/puts) to monetize elevated NVDA vols. Size positions to 1–3% of portfolio per idea and scale over 2–8 weeks; exit/trim on clear catalysts (GOOGL +20% or NVDA reclaiming prior ATH with volume).

Contrarian angles: Consensus underestimates durability of Nvidia’s software/hardware ecosystem (CUDA, models, developer inertia) — TPUs may complement rather than replace GPUs for many workloads. Valuation risk: GOOGL’s recent rerating could be mean-reverting if cloud margins disappoint; crowded long-megacap trades increase beta and correlation risk. A regime where regulators limit data/model usage could flip winners into losers quickly, so hedge sizing matters.

Key Decisions for Investors

• Establish a 2–3% net-long exposure to Alphabet (GOOGL/GOOG) over the next 4 weeks by scaling 30/30/40 (weeks 1/2/3–4), preferring 12–18 month LEAP calls if capital-efficient; trim if GOOGL rallies >20% from entry or if next two quarterly results miss revenue or cloud growth targets by >200bp.
• Reduce Nvidia (NVDA) outright equity exposure by 20–30% within 2 weeks to lock gains; monetize by selling 30–60 day covered calls (sell-to-open) at ~5–10% OTM to generate yield while retaining upside, or sell outright into sizeable up-days (>3% intraday volume spike).
• Initiate a relative-value pair: long GOOGL (1–2% portfolio) vs short NVDA (1–2%) funded equally, implemented via buying GOOGL shares or LEAP calls and buying 3–6 month NVDA 10% OTM puts to cap tail risk; rebalance if spread narrows by 50% or widens beyond stress thresholds.
• Rotate 3–5% of sector weight into cloud/software exposure (MSFT, AMZN, selected SaaS names) funded by reducing semiconductor cyclical exposures (TSM/AVGO) by similar amount; re-evaluate after next two earnings cycles (60–90 days) for margin signal.
• Monitor two binary catalysts closely and act: (1) US/EU chip export policy updates — if restrictive measures expand within 60 days, reduce NVDA by an incremental 10–20%; (2) GCP contract disclosures or TPU benchmark wins — if confirmed enterprise multi-year deals >$500M emerge, add incremental 1% to GOOGL.