Warren Buffett Says He Now Likes “Four or Five” Businesses Berkshire Owns More than Alphabet. What Are They?

Warren Buffett said he doesn’t “like” Alphabet as much as “four or five” other Berkshire holdings, pointing to the AI arms race’s massive capital intensity—Alphabet guided 2026 capex at $175B to $185B and Q1 2026 capex hit $35.67B (more than double a year ago). The article contrasts this with Buffett-style businesses with steadier reinvestment needs and highlights recent strength in other Berkshire names, including American Express (Q1 EPS $4.28, revenue $18.91B, billed business $428B up 10%) and Moody’s (Q1 revenue $2.08B up 8.1% and raised full-year buyback guidance to ~$2.5B). Sentiment toward Alphabet is cautioned even as the stock is up 18.5% YTD and 102% over the past year.

Analysis

The market implication is less about Buffett’s taste and more about regime: AI leadership is shifting from software scarcity to balance-sheet intensity, which should widen dispersion between capital-light compounders and capex-heavy platforms. That is incrementally positive for MCO, KO and AXP, where pricing power and low reinvestment needs allow earnings to compound without a matching step-up in capital outlay. OXY benefits more indirectly as a capital-return story in a world where investors increasingly reward free cash flow over growth-at-any-price, while BRK.B functions as a low-drama quality proxy rather than a fresh catalyst.

Near term, this is mostly a factor-rotation setup, not a fundamental break in GOOGL. The first real test is the next capex and margin update: if management keeps lifting investment without a clear path to incremental monetization, the multiple can compress even if revenue still grows. Over 1-3 months, the trade is about whether the market starts to price in lower FCF conversion; over 6-18 months, the bigger risk is that hyperscaler capex crowds out returns elsewhere and forces a supply-chain reallocation toward power, networking, and infrastructure rather than the platforms themselves.

The contrarian view is that the consensus may be overreading capex as a permanent destroyer of value. If GOOGL’s search, cloud and enterprise AI products monetize faster than expected, the ‘capital intensity’ critique becomes a temporary headwind and not a structural short. That argues for restraint on outright short exposure unless the next guide shows another step-up in spending or a clear margin inflection lower.

Key Decisions for Investors

• Long MCO / short GOOGL as a 1-3 month pair trade. Thesis: MCO monetizes AI financing/issuance while GOOGL carries capex risk. Target 8-12% relative outperformance; exit if GOOGL next guide shows stable margins despite higher capex.
• Buy KO on any market weakness for a 6-12 month quality-compounder trade. Low reinvestment intensity should outperform if investors continue to favor durable cash generation. Falsify if volume decelerates meaningfully or FX becomes a larger-than-expected drag.
• Add OXY on pullbacks as a capital-return / deleveraging expression over 6-12 months. Risk/reward improves if the market keeps rewarding free cash flow and balance-sheet repair; stop if oil falls below a durable $65/bbl support zone or debt paydown slows.
• Avoid initiating BRK.B solely as a headline-driven sympathy trade. The stock is more likely to track broad quality-factor flows than generate alpha from this news alone unless we see a sustained rotation into value-compounders.