GOOG · Alphabet Inc. · Stock research

ps

GOOG — Alphabet Inc. (Class C)

Current Thesis

The structural line this coverage flagged in late July — the ~$370 shelf — broke on the Q2 print and has not been reclaimed. Alphabet reported 2026-07-22: revenue $119.8B, +24% YoY, against roughly $116.5B consensus; Cloud $24.8B, +82% YoY with backlog at $514B. The stock fell anyway, -3.65% in after-hours trade to $329.43, because 2026 capex guidance went to $195–205B from $180–190B. That is the whole argument in one session: the operating business is accelerating and the tape is discounting the spend. The leg an investor buys here is that a $514B Cloud backlog eventually converts the capex into earnings before the market's patience runs out. Nothing dated resolves that inside 30 days — the next binary is the Q3 report in late October.

Bullish and bearish views on Alphabet Inc.

The model's bull view on Alphabet Inc. (GOOG), in brief: Q2 2026 (2026-07-22): revenue $119.8B, +24% YoY; operating income $40.8B, +30%; operating margin 34.0% vs 32.4% a year earlier. The bear view: The spend is being punished, not rewarded. Both cases follow in full.

Bull Case

• Q2 2026 (2026-07-22): revenue $119.8B, +24% YoY; operating income $40.8B, +30%; operating margin 34.0% vs 32.4% a year earlier.
• Cloud $24.8B, +82% YoY, up from +63% in Q1; segment operating margin 35.6% vs 20.7%; backlog $514B vs roughly $460B disclosed at Q1.
• Token processing 22B/minute in Q2, up from 16B (Q2 slides, 2026-07-22) — a demand metric management chose to put on the deck.
• Search & Other $63.3B, +17% YoY (2026-07-22) — the disintermediation thesis still has no printed evidence behind it.
• Waymo cleared by the CPUC on 2026-08-14 to expand autonomous ride-hailing across the SF Bay Area and Los Angeles and to enter Sacramento — regulatory access to California's largest metros, per Waymo's own post.
• Berkshire Hathaway raised its holding to 27.2M Class C shares; Lone Pine +14.9% to 215,574 Class A; Appaloosa +7% to 1,850,000 (13Fs filed 2026-08-14, positions as of 2026-06-30).
• Private-stake marks: Alphabet carries equity in Anthropic and SpaceX. Anthropic's Q2 revenue exceeded $11.5B (Bloomberg, 2026-08-14), and bankers were reported on 2026-08-15 framing $47B to $200B by 2028 ahead of an IPO — that repricing runs through Alphabet's balance sheet.

Bear Case

• The spend is being punished, not rewarded. Capex guide raised to $195–205B on 2026-07-22; shares dropped in the immediate reaction despite the revenue beat.
• Adjusted EPS $2.85 missed the $2.89 consensus while GAAP EPS of $9.11 was inflated by unrealized gains on the SpaceX and Anthropic stakes (Q2 recap coverage, 2026-07-22) — the headline number is not the operating number.
• Free-cash-flow framing has turned against it: a 2026-08-14 analyst note using MarketScreener estimates has Microsoft leading Big Tech FCF in 2027 while GOOGL and META stay negative.
• Gemini 3.5 Pro still is not generally available. Announced at I/O on 2026-05-19 with a June GA target; June, July and a widely reported 07-17 date all passed, and as of 2026-08-08 it remained in limited Vertex AI preview. The public flagship in the Gemini app is 3.5 Flash.
• Competitive prints keep landing: Anthropic Q2 revenue >$11.5B (2026-08-14); Alibaba AI model downloads reported passing Meta and Google (2026-08-16).
• Institutional distribution in Q2: Pershing Square exited Alphabet entirely; Viking Global dissolved its Class A stake; Coatue cut 18.6% to 908,395 Class C (all 13Fs filed 2026-08-14, as of 2026-06-30).
• Balance-sheet claims to verify: post-print recaps (Tickeron Q2 recap, 2026-07-22) reported long-term debt near $98B and buybacks suspended. Neither appears in the slide summary; both need checking against the 10-Q before being leaned on.
• Issuance into strength: the ~$85B equity raise priced 2026-06-03, upsized from $80B, funds the build but arrives ahead of the returns.

Setup & Price Structure

Last close $343.54 (2026-08-14). The 52-week high is $398.80, leaving price 13.9% below it; three-month return -12.6%; RSI(14) 59.2. Price sits between two dated references: the 2026-07-22 post-print reaction near $329.43 and the ~$370 shelf lost on that print. An RSI near 59 with price still 14% under the high describes a recovery that has retraced part of the drop without reclaiming any level that mattered. Overhead is $370, then $398.80. Volume-confirmed weekly acceptance back above $370 would be the first evidence a base exists; short of that, the $329–$370 band is where the capex argument gets settled.

The narrative is maturing. Operating momentum is still expanding (Cloud +82%, backlog $514B, 2026-07-22) and fresh dated headlines are landing (CPUC approval 2026-08-14), but participation is rotating rather than widening — four large managers cut or exited in the June quarter per 13Fs filed 2026-08-14, and the dominant frame since 2026-08-14 is which hyperscaler is FCF-positive in 2027 rather than which one spends most. It flips to saturated on a rejected retest of $370 with that FCF framing intact.

Crowding observables, without a verdict attached: no earnings inside the window removes the near-term binary; the ~$85B June raise and the reported buyback suspension both subtract corporate bid if confirmed; the 13F cross-section is genuinely two-sided (Berkshire, Lone Pine and Appaloosa adding against Pershing, Viking, Tiger and Coatue cutting); and the name sits below, not extended above, its July moving-average structure.

Catalyst Calendar (next 30 days)

• ~2026-10-27 (est., outside the window): Q3 2026 report — the next dated binary, carrying the 2026 capex guide and the Cloud growth rate.

Elapsed catalysts

• ~2026-08/09 (est., unconfirmed): Gemini 3.5 Pro general availability. Four target dates have passed since the 2026-05-19 I/O announcement; a fifth slip is itself a datapoint. (passed 99d ago)
• Undated, following 2026-08-14: Waymo commercial launch in Sacramento and expanded Bay Area / LA service under the CPUC approval. No start date has been published. (passed 12d ago)
• No confirmed Alphabet-specific dated event falls inside the window as of 2026-08-16. The DC Circuit appeal of the search monopoly ruling and the DOJ cross-appeal have no scheduled date in the period. (passed 10d ago)

What Would Change Our Mind

The break that matters already happened: $370 gave way on 2026-07-22 and every bar since has been a bounce beneath it. What would confirm that bounce as a lower high is a weekly close below $328, which surrenders the post-print reaction zone and puts the July low back in play as the reference. Second, a fifth Gemini 3.5 Pro slip — no general availability by 2026-09-30 — would take the frontier-parity leg off the table entirely rather than deferring it. Third, a Q3 print in late October that lifts 2026 capex above $205B while Cloud growth decelerates out of the 80s would confirm the spend-versus-return complaint that drove the July selloff.

In the other direction: a weekly close back above $370 on expanding volume, with the Cloud backlog growing again at the Q3 print, would re-establish the structure this coverage lost in July and change the read on the current range.

Correlation Notes

• Trades as a member of the megacap-AI cohort; the 2026-08-14 FCF note ranks it directly against MSFT and META, and headlines about one increasingly reprice all three.
• Dominant weight in the communication-services complex — XLC was flagged 2026-08-14 as a "no hike" scenario winner, so the name carries rate-path beta independent of its own fundamentals.
• Two-sided exposure to Anthropic: frontier releases and revenue prints (>$11.5B in Q2, 2026-08-14) pressure Gemini's competitive position while lifting the marks on Alphabet's equity stake.
• Inverse to the AI-capex complex on its own news — a raised capex guide is a demand event for semis, power and data-center names and a de-rating event for Alphabet, which is exactly how 2026-07-22 traded.
• Waymo headlines correlate to the robotaxi narrative broadly; the 2026-08-14 CPUC approval landed the same week as Tesla/SpaceX coverage dominated the autonomy tape.