3 Stocks Berkshire Hathaway Owns That I'd Buy Right Now, Including One It's Selling

Berkshire Hathaway's portfolio commentary highlights American Express, Alphabet and Ally Financial as attractive opportunities, despite Berkshire trimming its Ally stake by about 7% in Q2. Alphabet's Google Cloud posted 82% quarterly revenue growth and has a $514 billion backlog, supporting the case for its AI exposure and relative valuation. Ally, which has more than $100 billion in retail deposits, trades below book value and could benefit from wider net interest margins in a higher-for-longer rate environment, although it carries greater consumer-credit sensitivity.

Analysis

This is low-information retail commentary rather than a new fundamental catalyst; Berkshire ownership changes should not be treated as directional signals without reconciling the relevant 13F, ownership-percent calculations, and subsequent repurchase activity. The only potentially tradable implication is relative: a small reduction near a regulatory ownership threshold is mechanically different from a risk-driven exit, so ALLY should not inherit the negative read-through attached to larger bank disposals. That distinction is likely already understood by specialist financials investors, limiting immediate alpha.

GOOG remains the highest-quality expression of the group, but the key debate is not headline cloud growth: it is whether AI infrastructure depreciation and inference costs absorb incremental gross profit faster than monetization scales. Over the next 1-3 quarters, ad pricing, Search query-share data, Cloud operating-margin progression, and capex guidance matter more than backlog claims. A sustained capex acceleration without corresponding Services or Cloud margin leverage would compress the valuation premium and favor MSFT; evidence of stable Search economics plus margin expansion would support multiple expansion.

AXP is more exposed to a slowdown in discretionary travel, entertainment, and small-business spend than its affluent-customer framing implies, but its closed-loop model gives it superior visibility into payment volume and credit behavior versus network-only peers. ALLY offers greater upside if funding costs decline, yet it is a convex bet on used-car values, loss severity, and deposit retention; a modest NIM benefit can be overwhelmed by higher provisions. The contrarian view is that the apparently cheapest bank may remain cheap until auto-credit normalization is visible in net charge-offs and reserve releases, not merely because rates begin falling.

Near term, avoid chasing a sentiment move from this article. The 6-18 month structural winner is likely GOOG if AI monetization becomes measurable; the main second-order loser could be lower-scale software and internet firms facing higher AI compute costs without comparable distribution or advertising data advantages.

Key Decisions for Investors

• Maintain/establish long GOOG versus short MSFT in equal dollar size over a 3-6 month horizon only if GOOG's next earnings show Cloud margin expansion alongside contained capex growth. Target 10-15% relative upside; exit if Cloud profitability deteriorates materially or management raises annual capex without a monetization bridge.
• Use ALLY as a watch-list long, not an immediate recommendation: initiate only after quarterly net charge-offs and used-vehicle residual-value trends stabilize for two consecutive reporting periods. The upside case is book-value discount closure and lower funding costs; invalidate below a 10% deterioration in tangible book value or a material upward revision to credit-loss guidance.
• Prefer AXP over COF as a 6-12 month consumer-finance quality pair, sized modestly. AXP should outperform if premium spend remains resilient while COF remains more sensitive to normalization in card losses; close the pair if AXP billed-business growth decelerates below COF's for two consecutive quarters or AXP reserve builds accelerate.
• Do not trade BRK.A/BRK.B on the reported portfolio activity. Confirmed changes in Berkshire stakes are generally lagging disclosures, and the portfolio is too diversified for these positions to alter Berkshire's earnings power or near-term valuation.