Is This the Undiscussed Reason Buffett Just Bought Alphabet (Google) Stock?

Berkshire Hathaway trimmed its Apple position by roughly 41.8 million shares (about 14.9% of its Q2 position), approximately $9.4 billion of sales at period-average prices, while the remaining stake still increased in value by about $3.2 billion during the quarter; the firm deployed roughly $4.3 billion into Alphabet. Separately, Apple agreed to pay Alphabet about $1 billion annually to run a bespoke Gemini LLM on private Apple servers to power Siri—a strategic AI tie-up—while Apple remains Berkshire's largest equity holding (down from ~50% of the portfolio at end‑2023 to just over 21%), trades at a P/E near 37, and is benefiting from strong iPhone (including iPhone 17 in China) demand.

Analysis

Market structure: Berkshire's trimming of AAPL (~41.8M shares ≈ $9.4B sold) and a $4.3B GOOG buy plus a reported $1B/yr Gemini licensing deal creates a clear two‑way flow: GOOG gains recurring revenue and pricing leverage for LLMs while AAPL shores up product stickiness (Siri) without building public LLM hosting. Expect incremental demand for data‑center GPUs (NVDA) and private cloud infra; Apple services margin could rise modestly (+$1B/yr is ~0.3% of AAPL 2024 revenue) while competitive pressure on Amazon/Alexa for assistant capability increases.

Risk assessment: Tail risks include antitrust/regulatory scrutiny of a revenue-sharing Google‑Apple pact, a model/data breach causing reputational/legal losses, or supply crunch in GPUs that spikes costs 20–40% short term. Immediate (days) — headline-driven volatility in AAPL/GOOG; short term (weeks–months) — re‑rating as Qs incorporate licensing revenue; long term (quarters–years) — structural shift in assistant monetization and services stickiness. Hidden dependency: Apple’s reliance on Google for core LLM tech creates strategic optionality loss and potential margin leakage.

Trade implications: Direct plays — overweight GOOG and NVDA (AI infra), trim AAPL concentration or hedge it. Pair trades — long GOOG / short AAPL on a 3–12 month horizon to express monetization vs hardware growth divergence. Options — buy 3–9 month GOOG calls to capture re‑rating, and buy 6‑month AAPL puts (5%–10% OTM) as insurance; size per bullets below.

Contrarian angles: Market may mistake Berkshire rebalancing for loss of conviction in AAPL — this is likely risk management; GOOG’s monetization is underpriced relative to historical MSFT+OpenAI precedents where licensing scaled over multiple years. Unintended consequence: Apple outsourcing core AI may slow its internal LLM roadmap and create longer‑term competitive risk; monitor Siri usage and services revenue as early signals.

Key Decisions for Investors

• Establish a 0.75%–1.5% core long position in Alphabet (GOOG) within 2 weeks, target 12–18 months; cut position if price drops >12% from entry or take profits at +25–30% or after two quarters of confirmed LLM revenue recognition.
• Add a 0.5%–1.0% tactical long in NVIDIA (NVDA) within 1 month to play GPU demand from LLM hosting; use a 15% trailing stop and consider selling into a 40% rally or if data‑center revenue growth slows below 20% YoY on two consecutive quarters.
• If AAPL weighting >10% of portfolio, trim exposure by 3%–5% (rebalance into GOOG/NVDA). If AAPL position is smaller, buy 6‑month AAPL puts (5%–10% OTM) sized to insure 50% of the position cost to protect versus a >10% drawdown.
• Implement a dollar‑neutral pair trade: +1% GOOG / −1% AAPL for 3–9 months to express AI monetization tailwinds; unwind or rebalance if spread moves >15% or after the next two earnings releases confirm revenue trajectories.
• Monitor four catalysts over next 30–90 days: (1) Apple/Google revenue recognition details and service‑revenue impact, (2) Berkshire 13F changes, (3) NVDA supply commentary, (4) WWDC/Apple product roadmap; convert hedges to cash if Apple reports >$1B incremental services by FY+1 or if regulatory investigations are announced.