Berkshire Is Selling Apple Stock and Buying This Other Magnificent Artificial Intelligence (AI) Stock Instead

Berkshire Hathaway materially pared its Apple position from 906 million shares in Q4 2023 to 238 million in Q3 2025 (roughly a 73% reduction) while accumulating cash and interest in Treasury bills and deploying some excess capital last quarter to acquire a stake in Alphabet. The article highlights Alphabet's AI investments — Gemini integration, custom TPUs, and a $4.7 billion Intersect acquisition — and notes a forward P/E of about 29 alongside fresh purchases by major investors (Druckenmiller, Englander, Griffin, Laffont and Buffett). It frames Berkshire's Apple sales as macro-driven profit-taking after ~60% market gains since October 2023 and positions Alphabet as a compelling long-term AI-era play.

Analysis

Market structure: Berkshire’s rotation out of AAPL (≈73% reduction since end-2023) and into GOOGL signals a flow shift from mature consumer hardware toward AI-enabled ad/cloud leaders. Winners: GOOGL, NVDA, AMZN, MSFT (AI infra, cloud, ad monetization); losers: hardware-dependent AAPL and cyclical consumer names if rates reprice. Supply/demand: rising demand for TPUs/GPUs and data‑centre power points to sustained semiconductor and power-infrastructure capex; Treasury bill yields remain an alternative cash sink, muting some equity upside.

Risk assessment: Key tail risks are antitrust action versus Alphabet, catastrophic AI product failures (monetization delays for Gemini), and a sharp AI-hype drawdown (30–40% tech correction). Immediate (days) risks are momentum and options vol; short-term (weeks/months) hinge on earnings/ad-revenue prints and NVDA guidance; long-term (years) depend on TPU/cloud adoption and regulatory outcomes. Hidden dependency: Alphabet’s margin leverage to ad recovery and energy costs (Intersect acquisition reduces long-term power risk but raises short-term capex).

Trade implications: Favor concentrated positions in GOOGL (core long) and NVDA (infra exposure) with size caps: 2–4% and 1–2% portfolio respectively; trim AAPL exposure materially and hedge with put spreads. Use 9–18 month call spreads on GOOGL to capture valuation expansion while limiting premium, and buy short-dated protective puts on NVDA into earnings. Rotate sector weight from consumer hardware into AI infra/cloud over next 2–6 weeks and set profit trims at +25–35% or P/E thresholds (GOOGL >35).

Contrarian angles: Consensus underestimates regulatory/legal execution risk for GOOGL and overestimates near-term monetization of Gemini; Apple may be oversold relative to free-cash-flow yield and buyback muscle. Historical parallel: 2016–18 FAANG rotations showed megacap leadership can reassert after a pullback—watch ad revenue inflection as the deciding signal. Unintended consequence: concentration into AI names increases systemic risk if GPU shortages or Fed tightening re‑prices growth multiples.

Key Decisions for Investors

• Establish a 2–4% net long position in GOOGL within 2–6 weeks: 60% equity, 40% 12–18 month call spread (buy LEAP ~Jan 2028 OTM call, sell a higher strike to offset cost). Trim by 25% if forward P/E > 35 or price rises >30% in 6 months.
• Allocate 1–2% to NVDA as core infra exposure: buy shares or 9–12 month call options sized to emulate shares; if implied vol >50% ahead of earnings, hedge with 4–6 week 7–10% OTM puts. Take profits if position gains >30% or company lowers Q guidance.
• Reduce AAPL exposure by 50% (or to ≤1% portfolio) within 30 days; if retaining any exposure, purchase a 3–6 month put spread (e.g., 5–10% OTM) sized to cover the residual holding to protect against a 20% downside in the next 3 months.
• Execute a dollar-neutral pair trade: long GOOGL / short AAPL sized 0.8:1 (long leg smaller) for a 6–12 month horizon. Exit if the share-price spread narrows by 20% or if Alphabet misses ad revenue by >3% vs consensus or faces a major regulatory filing.