Warren Buffett Bought Alphabet Stock About a Year Ago. 3 Reasons It’s Still a Buy Today

Alphabet reported Q2 revenue of $119.8 billion, up 24% year over year, while Google Cloud revenue surged 82% to $24.8 billion and its backlog reached $514 billion. The company raised 2026 capital-expenditure guidance to $195 billion-$205 billion from $180 billion-$190 billion, signaling continued AI infrastructure investment. A federal judge rejected a U.S. government effort to force the sale of AdX, reducing a major antitrust downside risk, though further legal exposure remains. Berkshire Hathaway has initiated and subsequently increased its Alphabet position to one of its five largest holdings.

Analysis

The investable issue is whether Alphabet can convert AI infrastructure spend into durable cloud gross-profit dollars before depreciation and power costs pressure consolidated margins. A growing contracted-revenue base improves utilization visibility, but it does not establish return on invested capital: monitor cloud operating-margin progression, capex-to-revenue intensity, and the share of backlog that converts to recognized revenue over the next 2-4 quarters. If cloud growth remains elevated while segment margins hold or expand, Alphabet merits multiple expansion versus slower-growth ad-platform peers; if monetization lags capacity additions, the stock becomes a capex-cycle de-rating candidate.

The less appreciated beneficiary is the AI infrastructure supply chain—NVDA, Broadcom (AVGO), Arista (ANET), and data-center power/cooling vendors—because Alphabet's spending plans support orders before its own cloud revenue is realized. Conversely, aggressive pricing and bundled AI services could pressure AWS and Azure unit economics, making GOOG potentially preferable to MSFT on incremental cloud-growth exposure, although Microsoft has better enterprise software monetization offsets. Over 6-18 months, Waymo is strategically valuable primarily as an option on autonomy; it should not drive near-term valuation absent disclosed ride volume, unit economics, or external funding marks.

The antitrust risk is no longer binary breakup risk alone: behavioral remedies, default-distribution restrictions, data-sharing requirements, or ad-tech interoperability mandates could reduce search monetization and raise traffic-acquisition costs without requiring a divestiture. Consensus may be too sanguine on this earnings risk and too focused on favorable procedural developments; a remedy that lowers search query share or ad yield would matter more than a one-time legal charge. Near-term news impact is likely limited unless a remedy timetable, appeal ruling, or a change in capital-return policy creates a concrete earnings-estimate revision.

Key Decisions for Investors

• Accumulate GOOG on market or AI-capex-driven pullbacks for a 6-18 month horizon, but size as a core long rather than a catalyst trade. Thesis requires cloud segment margin stability and no material reduction in consolidated operating-margin guidance; cut exposure if either deteriorates for two consecutive quarters.
• Run a 1-3 month relative-value watch: long GOOG / short MSFT only after next earnings confirm superior cloud growth without incremental margin dilution. Target 8-12% relative upside; invalidate if Azure reaccelerates materially or Alphabet raises capex again without matching backlog conversion or margin support.
• Maintain NVDA/AVGO/ANET exposure as second-order beneficiaries of hyperscaler deployment, but use staggered entries rather than chasing announcement-driven strength. The key risk is a shift from capacity buildout to utilization optimization; reduce if hyperscaler capex guidance flattens or lead times normalize.
• Do not underwrite Waymo in GOOG valuation yet. Upgrade its contribution only when Alphabet discloses repeatable city-level utilization, contribution margins, or a third-party financing valuation that independently validates economics.
• Set a regulatory alert around remedy proposals and appellate milestones: hedge GOOG event risk with limited-duration put spreads only when implied volatility is below prior legal-event ranges, rather than carrying costly permanent protection.