Google Stock Drop — Cloud +82%, Capex Guide to $205B

The twist: this was not a failed quarter. Revenue hit $119.8B (+24%), beating estimates, and Google Cloud grew +82%, well ahead of the Street. Shares still sold off because investors focused on Capex guidance and the quality of earnings more than the top line.

Below we unpack the segment numbers, why $9.11 EPS is misleading, what Capex and free cash flow (FCF) are saying, and what to watch into Q3 and 2027. This is not a buy/sell recommendation for any stock or ETF.

At a glance — good news and bad news in one print

Capex is cash spent on long-lived assets such as data centers and servers. FCF is operating cash after that spending—what is left (or short). Alphabet still generated large operating cash, but investment outran it, so FCF turned negative for the quarter.

• Positives: Revenue $119.8B beat; Cloud $24.8B (+82%); cloud operating margin 35.6%; backlog $514B
• Optical illusion: Diluted EPS $9.11 — mostly equity-mark gains (~$99B). Core-like EPS ~$2.85 (vs ~$2.89 expected; slight miss)
• What the market hated: FY2026 Capex raised to $195–205B; Q2 Capex $44.9B (+100%); FCF −$5.9B
• Near-term worries: More third-party compute in Q3 → modest margin pressure; Search faces tough year-ago comps

1. 1
   ① Revenue & Cloud beat

   $119.8B / Cloud +82%

2. 2
   ② EPS optics

   ~$6.26 of $9.11 from marks

3. 3
   ③ Capex raised

   FY guide $195–205B

4. 4
   ④ Negative FCF & AH drop

   −$5.9B; shares ~−4% AH

Q2 numbers — revenue and operating profit first

Consolidated revenue was $119.8B, up 24% year over year (23% in constant currency), beating LSEG’s ~$116.93B. It was Alphabet’s 12th straight quarter of double-digit revenue growth.

Core profitability was solid: operating income $40.8B and a 34% operating margin (+2 percentage points). The business is not ‘broken.’ The issue sits lower on the income statement—net income and EPS—where equity valuation gains loom large.

Reported diluted EPS was $9.11 (+294%) and net income available to common was $112.1B. Other income was about $98.0B, driven by ~$99.0B of gains on equity securities. Per the 8-K FAQ, those gains added about $6.26 to EPS—implying a core-like EPS near $2.85. CNBC’s adjusted EPS of $2.85 also slightly missed LSEG’s $2.89.

In plain terms, much of the profit was not cash that hit the bank—it was a mark-up on stakes such as Anthropic and SpaceX. If those marks reverse next quarter, the headline swings the other way. Reports that SpaceX shares softened after quarter-end fed skepticism about how ‘durable’ that EPS is.

Segments — Cloud stole the show; Search was tight

Google Services revenue was $94.5B (+15%). Search was $63.3B (+17%), a slight miss versus StreetAccount’s ~$63.4B. YouTube ads were about $11.06B (+13%), ahead of estimates in the live coverage.

The star was Google Cloud: $24.8B of revenue, +82% year over year, versus expectations near +64% / ~$22.4B. Cloud operating income hit $8.8B, more than tripling, with margin rising from 20.7% to 35.6%. Backlog reached $514B, up more than $50B sequentially.

Growth came from GCP enterprise AI solutions and infrastructure plus core GCP. Alphabet also recognized TPU system sales delivered into customer data centers for the first time; management said growth still accelerated excluding that. Other Bets (including Waymo) contributed $382M (+2.4%)—still tiny versus the core.

Why it fell — Capex $195–205B and negative FCF

Q2 Capex was $44.9B, up 100% year over year. CFO Anat Ashkenazi said most of it went to technical infrastructure for AI—about 60% servers and 40% data centers and networking.

The bigger shock was the full-year guide: Capex of $195–205B, up from $180–190B last quarter and above the ~$188B the market had been modeling. Management framed it as accelerating capacity for demand. Equity markets heard more cash leaving the building.

That shows up in FCF. Non-GAAP free cash flow was −$5.855B: $39.069B of operating cash versus $44.924B of property and equipment purchases. Spending briefly outran cash generation—common in AI buildouts, but hard on multiples when the stock already priced in a lot of success.

Q3 color added fuel. In a supply-constrained environment, Alphabet plans to use more third-party capacity as a bridge, which management said will create modest near-term margin pressure. After-hours pops in names like CoreWeave and Nebius showed how fast the tape priced ‘Google will rent more outside GPUs.’

AI, ads, Search — product signals

CEO Sundar Pichai said AI investments are redefining what is possible across the business. The Gemini app has about 950 million monthly active users and processes about 22 billion API tokens per minute (up from 16 billion). Antigravity, the coding tool, has 2.4 million+ weekly active users.

On ads, customers using tools such as Performance Max and AI Max saw about 50% more conversions on Search at similar ROAS, the company said. During the FIFA 2026 World Cup, Search usage hit an all-time high and YouTube logged 1.7 billion unique viewers of World Cup–related video.

Search comps get harder in Q3 as Alphabet begins lapping last year’s acceleration. The CFO said AI continues to improve the user and advertiser experience—but investors should expect tougher year-over-year compares. When Cloud is this strong, the tape gets pickier about a tiny Search miss and the next quarter’s path.

Outlook — four things to watch

This is a watchlist, not a price target. The broader AI infrastructure cycle overlaps our semiconductor selloff and bottleneck outlook.

1. Backlog → revenue: How fast the $514B Cloud backlog converts into sales and margin
2. Capex & FCF: Pace of the $195–205B spend and when quarterly FCF turns positive again. 2027 was already flagged to ‘significantly increase’ (April call); street chatter has cited mid/high-$250B estimates
3. Search growth: Whether mid-teens growth holds through the Q3 lap, and whether AI Mode defends queries and ad pricing
4. Earnings quality: If Other income / equity marks swing again, headline EPS will distort—pair it with operating EPS and margin

Pre-print aggregations showed a heavy Buy skew and average 12-month targets near $430. That was a pre-earnings snapshot; Capex resets force target and multiple revisions. For index vs dollar regimes, see our US market correction and DXY history note.

The bull path is straightforward: Cloud keeps compounding, Capex unlocks capacity, and margins re-expand. The bear path is Capex staying ahead of cash generation, Search decelerating, and mark-to-market noise dominating the P&L. Both are live paths—neither is proven yet.

Checklist for Korea-based readers

Many Korean investors hold GOOGL/GOOG directly or via US big-tech / Nasdaq ETFs. For ETF fees and tax structure, see our Korea–global ETF tax and fee guide. Liquidity and rates context also overlaps Fed hold / wallet impact.

• Do not stop at headline EPS ($9.11)—check core EPS and operating margin
• Read Cloud growth together with Capex guidance and FCF
• Next print: Q3 Search growth, third-party capacity costs, and margin commentary
• In ETF baskets, watch Google’s weight plus co-movement with other AI Capex-sensitive names

Bottom line

This Google stock drop was less ‘bad earnings’ than an A-grade Cloud report with a bigger Capex invoice. $119.8B revenue and Cloud +82% are real positives; $9.11 EPS was inflated by equity marks; $195–205B Capex and negative FCF tripped the stock. The next test is how fast backlog becomes profit.

This summary is based on filings and press coverage and is not investment advice. Trading and sizing decisions should rest on your own judgment and the latest disclosures.

---

Cloud scored an A; Capex was the invoice the tape priced first.

Sources

• CNBC — Google Q2 2026 earnings live updates
• Alphabet 8-K / StockTitan filing summary
• Investing.com — earnings call transcript summary
• MarketBeat — GOOG Q2 2026 earnings / transcript
• TradingKey — selloff after EPS beat
• InsiderFinance — preview & analyst targets
• Internal: 2026-07-23-alphabet-q2-earnings-stock-drop-session-01.md