GOOGL: FY25 Deep Dive

Capex jumped 74% to $91B. Cloud operating margin crossed 23%. FY25 is the year the AI-infra spend inflection showed up in both the numerator and the denominator — with $37B of new debt funding the delta.

Key Takeaways

Alphabet closed fiscal 2025 (calendar year ended December 31, 2025) at $402.8 billion of revenue, up 15% year-over-year. Google Services still carries 85% of the top line, but the structural shift this year is Google Cloud — revenue grew 36% to $58.7 billion and segment operating margin inflected from 14.1% to 23.7%, the cleanest year-over-year profitability step in any Alphabet segment since Search matured in the early 2000s. The AI-infra buildout showed up in both directions: capital expenditures stepped up 74% to $91.4 billion, long-term debt tripled from $12 billion to $49 billion to fund the step-up, and free cash flow held flat at $73.3 billion against $73 billion in FY24 despite revenue growing $52 billion. Buybacks fell $16.5 billion to $45.7 billion as cash was redirected into infrastructure, while dividends rose 36% to $10.0 billion. Sell-side sits at 83% Buy or Strong Buy across 23 analysts, median price target $386.30, range $310-$420.

Main business structure

Three reporting segments (Google Services, Google Cloud, Other Bets) plus a Hedging line; Google Services is internally broken into an Advertising bloc and Subscriptions/Platforms/Devices.

Segment profitability (operating income, FY25):

Consolidated operating margin held at 32% — the Cloud margin improvement offsets the rise in Other Bets and Alphabet-level losses (both widened, in part from AI research spending categorized centrally).

Geographic mix. Revenue is reported by billing address. US ~50%, EMEA ~30%, APAC ~14%, Other Americas ~6% — proportions roughly stable year-over-year. No single customer accounts for 10% of revenue disclosure is not triggered (the ad base is fragmented across millions of advertisers; Cloud has no single disclosed concentration).

Scale anchors. Search handles >90% of global query share outside China. YouTube monthly active users >2.5 billion. Android active devices ~3 billion. Google Workspace users ~3 billion. TPU v6 / v7 is Alphabet's in-house AI silicon, increasingly pitched externally as a Nvidia alternative. Gemini is the consumer-facing LLM product line; API / Vertex AI Cloud pricing is what shows up in the Google Cloud segment revenue.

Key core metrics (3-year+ trend)

Four metrics carry the story of Alphabet through FY25.

1. Google Cloud operating margin (the inflection)

A four-year linear march from loss-making to 24% operating margin. The 10-K language attributes the FY25 step to revenue growth outrunning the increase in technical-infrastructure usage costs and headcount — in other words, scale economies on a $91B capex base are reaching the income statement. This is the single most consequential trajectory in the Alphabet complex.

2. Capex ($B)

Capex intensity more than doubled in three years. The FY24→FY25 step ($52.5B → $91.4B, +$38.9B) coincides with the $37 billion in new long-term debt issued during the year — this is not a coincidence, it's how the AI-infrastructure buildout is being funded.

3. Free cash flow ($B)

The counterintuitive print: FCF was flat year-over-year despite $52B of incremental revenue, because the capex step absorbed the full OCF expansion. FCF margin compressed 260bp — the cost of the AI capex cycle, made explicit.

4. Capital return split ($B)

Buybacks were cut 27% in FY25 as cash rotated into capex; dividends (instituted only in FY24) grew 36%. Total capital return dropped $14B year-over-year. The capital-return program is no longer growing in step with earnings — the company has explicitly prioritized AI infra over incremental buybacks.

Market evaluation

Sell-side coverage (as of late April 2026). 23 analysts cover the stock.

Price targets. Median $386.30, range $310 (low) to $420 (high: BMO Capital).

Recent analyst activity (post-Q4 FY25 earnings in early February through April 24, 2026). Wells Fargo (Ken Gawrelski) upgraded the stock from Equal-Weight to Overweight on February 23 with a PT raise from $354 to $387. Six weeks later, April 2, the same analyst cut the PT by $36 to $361 — the largest single-move PT reduction across the 14 covered events in the window. UBS delivered the largest post-Q1 earnings raise on April 21 (+$27 to $375, Neutral maintained). BMO raised to $410 on April 22, the highest PT currently on the Street. The Hold camp (4 analysts, including UBS and Rosenblatt) cites the capex trajectory and the FCF compression as the reason to wait for a cleaner cash-flow inflection before re-rating.

Buy-side positioning. GOOGL is an overweight position in the vast majority of institutional technology portfolios, with no notable crowding signals at either extreme of the tape. Short interest remains below 1% of float.

FY25 corporate structure: debt issuance and Cloud re-rate

FY25 contains two structurally important corporate events that the 10-K documents explicitly. First, the long-term debt balance expanded from $12 billion at end of FY24 to $49 billion at end of FY25 — $37 billion of net new issuance, primarily in investment-grade unsecured notes across tenors, with proceeds disclosed as general corporate purposes and infrastructure spending. Alphabet was effectively debt-free through its history; FY25 changes that permanently. Second, Google Cloud crossed an internal threshold worth naming: operating income of $13.9 billion is now larger in absolute dollars than Other Bets' cumulative losses since 2016, and Cloud's 24% operating margin puts it in a different profitability regime than any sell-side model carried into FY25. Whether the Cloud margin expansion continues at the FY24-to-FY25 pace (+9.6pp) or normalizes to the FY23-to-FY24 pace (+8.9pp) is the single line in the Alphabet model that the largest number of Street analysts are actively debating — and the dispersion in FY26 Cloud margin assumptions is the main source of the $110 price target spread ($310 to $420).