Why we're sticking with Alphabet despite an imperfect quarter and more AI spending

Alphabet’s Q2 revenue rose 24% YoY to $119.8B (vs. $116.9B expected) and EPS was $9.11, but the bottom line is clouded by a ~$99B equity investment gain; excluding it, operational earnings were about $2.85/share vs. expectations, driving the stock down over 3% after-hours. Operating income beat at $40.77B and margins expanded 1.6pp, yet Google Search revenue missed and investors focused on higher capex as Alphabet raised 2026 spending guidance to $195B–$205B (from $180B–$190B). The company also flagged capex would “increase significantly” in 2027 and faces near-term supply overhang from a planned $40B at-the-market equity program starting in Q3.

Analysis

The market is likely punishing the mismatch between accelerating AI demand and deteriorating cash conversion. Near term, the stock trades less on reported operating income and more on whether incremental AI dollars are accretive enough to offset a higher capex path; that keeps the multiple vulnerable even if the business is still compounding. The key overhang is supply: when a mega-cap raises equity-linked liquidity and simultaneously tells investors capex will stay elevated, you create a rare setup where fundamental good news can be outweighed by balance-sheet and float concerns.

Second-order winners are the AI infrastructure names that monetize the spend without having to defend a consumer monetization story. AVGO is the cleaner beneficiary because custom silicon and co-design work should gain share as hyperscalers optimize cost per token, while the broader merchant GPU ecosystem could face periodic budget scrutiny as buyers diversify into in-house accelerators. The loser set extends beyond one name: any advertiser or software platform leaning on search-distribution traffic should watch for higher AI answer penetration to distort click economics over the next 1-3 quarters.

Contrarianly, the consensus may be too focused on near-term EPS optics and too dismissive of the cloud backlog and usage data, which usually lead reported AI monetization by several quarters. That means the fundamental thesis is not broken; the issue is timing. If capex grows faster than cloud revenue and search monetization for another quarter or two, the stock can de-rate again before the longer-term earnings power reasserts itself.

Key Decisions for Investors

• Pair trade: long AVGO / short GOOGL for 1-3 months. Thesis: AVGO captures AI infrastructure demand with cleaner cash conversion, while GOOGL remains exposed to capex fatigue and float/supply overhang. Risk/reward is attractive if the market keeps rewarding infrastructure monetization over spend-heavy platforms.
• Do not add aggressively to GOOGL on the first post-print bounce; wait for the ATM supply to start clearing and for the stock to show at least 2-3 sessions of stabilization. If it fails to hold the post-earnings low, expect another leg of multiple compression.
• Use any further 5-8% pullback in GOOGL as a staged long entry for a 6-18 month horizon only if management reiterates cloud backlog strength and next-quarter capex growth moderates. Falsifier: another capex raise without corresponding FCF improvement.
• Overweight AVGO versus the broader AI capex basket. If hyperscalers keep shifting toward custom silicon and lower-cost inference, AVGO should see incremental demand with less headline risk than the hyperscaler owners of that capex.
• Set an alert on GOOGL operating cash flow/free cash flow turning positive again. Until then, treat the name as a hold, not an add; a sustained negative FCF print would justify further de-rating.