31% of Berkshire Hathaway's Portfolio Is Riding on These 2 AI Stocks Under Greg Abel

Berkshire Hathaway has roughly 31% of its equity portfolio invested in Alphabet and Apple, including nearly 106 million Alphabet shares that lifted the position to just under 10% of the portfolio. Alphabet's revenue growth accelerated to 24% over the last year and Google Cloud growth reached 82%, while it generated $53 billion in free cash flow despite $195 billion-$205 billion of planned annual capex. Apple remains just over 20% of Berkshire's portfolio after significant trimming, supported by 16% fiscal Q3 2026 revenue growth and its Gemini-based AI partnership with Google, though its 38x P/E limits its appeal as a new purchase.

Analysis

The investable signal is not Berkshire’s AI “exposure” but the validation of Alphabet as a value-compatible compounder: a large, price-insensitive buyer can tighten the incremental float while reframing Google’s spend cycle as funded infrastructure rather than a cash-burn risk. That supports GOOGL multiple resilience over the next 1-3 months if Cloud backlog, search monetization, and operating-margin guidance continue to absorb elevated depreciation. The key falsifier is not headline capex, but a deceleration in Cloud growth or a material step-down in consolidated FCF conversion over the next two earnings reports.

AAPL’s Gemini dependency is more ambiguous than bullish. It can shorten time-to-market for consumer features and help defend upgrade demand, but it shifts a portion of the AI value pool toward the model/platform provider and reduces the likelihood that Apple earns an AI-specific valuation premium. Over 6-18 months, the relevant question is whether AI features raise iPhone replacement rates and Services ARPU enough to offset any gross-margin drag from inference costs, licensing, or subsidized hardware promotions.

BRK.A/BRK.B should not be treated as a clean AI proxy: its public-equity portfolio is economically meaningful but insufficient to drive the conglomerate’s near-term earnings multiple versus insurance underwriting, investment income, and capital deployment. Consensus may overread a reported stake change because ownership disclosures are backward-looking; the better near-term catalyst is confirmation of continued buying in the next filing, not the initial narrative. A broad AI risk-off move would likely hurt the look-through equity portfolio while simultaneously creating deployment opportunities for Berkshire’s balance sheet, limiting downside relative to a direct GOOGL position.

Key Decisions for Investors

• Initiate a 3-6 month pair: long GOOGL / short AAPL, sized beta-neutral. Thesis is that GOOGL captures both infrastructure and platform economics, while AAPL primarily captures uncertain device-cycle benefits; target 10-15% relative outperformance, cut if AAPL reports AI-led unit/replacement acceleration or GOOGL Cloud growth decelerates sharply.
• Add to GOOGL only on post-earnings volatility if management reaffirms investment spending while Cloud margin and consolidated FCF remain intact. Use a 12-18 month horizon; the primary risk is that depreciation and inference costs arrive ahead of monetization, producing an earnings-reset rather than a capex multiple expansion.
• Do not add BRK.A solely on this narrative. Use BRK.B as a defensive substitute for part of a megacap-tech allocation only if the next ownership filing confirms sustained purchases and insurance pricing remains favorable; otherwise, the direct GOOGL exposure offers cleaner upside.
• Watch AAPL’s next earnings for iPhone unit growth, gross-margin commentary, and Services ARPU rather than AI feature announcements. A measurable acceleration in these metrics would invalidate the short leg of the GOOGL/AAPL pair before any longer-term platform economics become visible.