Alphabet Stock Is Down 16% -- Is September the Month to Buy the Dip?

Alphabet shares are about 18% below their May peak as investors question whether roughly $200 billion of planned 2026 capital spending, largely for AI infrastructure, will generate adequate returns. The article argues Google is better positioned than peers: Google Cloud revenue grew 81% year over year last quarter and operating income more than tripled, supported by its integrated AI cloud stack, proprietary TPUs and specialized software offerings. It views the selloff as a buying opportunity despite broader uncertainty over AI-investment payoffs.

Analysis

The relevant debate is not whether Google Cloud can grow, but whether incremental AI revenue converts to durable returns above Alphabet's cost of capital before depreciation ramps. A $200B annual infrastructure program raises the hurdle materially: even strong cloud growth can be equity-negative if utilization lags build-out, enterprise workloads remain low-margin, or depreciation begins outrunning gross-profit expansion. The next 1-3 months should therefore be driven less by AI product announcements than by evidence on backlog conversion, Cloud operating-margin progression, and management's 2027 capex trajectory.

Alphabet has an underappreciated advantage versus MSFT, AMZN and AI model vendors: it can monetize compute internally through Search, YouTube, Workspace and ads, rather than relying exclusively on third-party cloud demand. That diversification lowers stranded-capacity risk, but it also makes ROI opaque; investors need proof that AI-driven query monetization and advertiser ROI offset any higher inference cost per search. If AI answers cannibalize paid-click inventory or query economics deteriorate, the premium attached to Google's distribution moat compresses quickly.

The consensus bullish framing may be early rather than wrong. TPU competitiveness is strategically valuable only if it creates a sustained unit-cost advantage versus NVDA-based infrastructure and translates into external Cloud share gains; it does not automatically justify a higher multiple. Near-term weakness could persist through the next earnings print if capex guidance rises again, while a credible capex plateau or accelerating Cloud margin would create a more investable rerating catalyst over 6-18 months.

Key Decisions for Investors

• Maintain a measured long GOOG/GOOGL only after confirming the next earnings release shows Cloud operating-margin expansion alongside no upward revision to forward capex; target a 6-12 month rerating, with thesis invalidated by decelerating Cloud growth plus another material capex-guide increase.
• Express relative preference through long GOOG / short AMZN in equal dollar exposure over 3-6 months: Alphabet's internal demand engines provide more paths to infrastructure monetization, while AWS is more exposed to enterprise-cloud pricing competition. Exit if AWS growth reaccelerates materially faster than Google Cloud or Alphabet's Cloud margin contracts.
• Do not chase NVDA from this development. A broader TPU deployment narrative is a medium-term competitive risk to accelerator pricing and could cap multiple expansion, but only becomes actionable if disclosed Google Cloud workloads or third-party TPU adoption demonstrably displace GPU demand.
• Set an event alert around quarterly depreciation, capex and Cloud-margin disclosures. A widening gap between capex growth and Cloud/advertising profit growth is the key signal to reduce GOOG exposure; a narrowing gap supports adding on post-earnings volatility.