The call: BUY GOOGL $315–$322 into Oct 28, target $355–$365, invalidation daily close <$295. Timeframe 2–4 weeks. Conviction: SPECULATIVE-HIGH.
Why now: GOOGL sits 21% below its May high while Google Cloud just grew 82%. Microsoft guides Azure ~45%. The market is pricing dilution and duration, not deceleration.
The disagreement: Consensus sees Warsh on hold as dovish. The CNBC Fed Survey shows 93% expect a pause this week but 76% expect a December hike — a hawkish hold that punishes unprofitable duration first, profitable cloud second.
The level that changes everything: 10Y at 4.99%. Below 4.80% the rally broadens. Above 5.10% halve all longs — my standing book rule.
October 28 delivers the Fed decision and the two most important AI earnings within hours of each other. The FOMC statement drops Wednesday afternoon, with Microsoft (2:30pm PT) and Alphabet (1:30pm PT) reporting the same day and Amazon following Oct 29, per the earnings calendar.
Stocks enter the gauntlet bid but fragile. As of ~16:00 UTC Oct 27: SPY $770.69 (+0.70%), QQQ $722.08 (+1.39%), MSFT $534.44, GOOGL $321.29 (+1.42%), META $626.35 (-0.67%), NVDA $218.88, BTC $80,506. Fear & Greed is 48 Neutral, down from 68 Greed on Oct 22. That sentiment reset is constructive — the market already de-risked.
Mike Santoli's now-or-never note for the AI trade frames it correctly: the S&P stabilized in a tight range, hedge funds cut tech, and retail dip-buying hasn't yet marked the low. What breaks the range is Wednesday. My opinion: this is not a normal earnings day — it is a joint monetary + micro solvency test. Position for the joint outcome, not one leg.
Hyperscaler revenue is the best it has been in two years — and cash conversion is the worst. That tension is the entire trade.
The numbers: Google Cloud grew 82% in Q2 to $24.8B, Azure grew 43%, AWS +37%, per Q2 recaps. Microsoft enters fiscal Q1 2027 with consensus Azure ~45% and capex $46.2B for the quarter alone — versus $19.4B a year ago. Morningstar into the print pegs fair value at $600 and calls MSFT moderately undervalued, flagging Azure growth, capex, and margins as the three variables.
Why it matters: every extra point of cloud growth now costs roughly $2B in quarterly capex. Alphabet already lifted 2026 capex guidance twice, to $195–$205B. Microsoft's Azure crossed $100B annual revenue but quarterly capex hit $41B and FCF fell 23% last quarter. The market no longer pays for growth alone — it pays for growth that still converts.
What I will watch line-by-line: Azure growth vs. 45% whisper, Google Cloud vs. ~70%+ needed to hold 80%+ momentum, and full-year capex language. A beat on cloud with flat-to-down capex guide is a 5–8% up day. A beat with a raise in capex is a sell-the-news day — exactly what happened to Alphabet in July when it beat and still fell.
Meta broke a 13-year silence on new stock because AI capacity is worth more diluted than delayed.
Meta plans up to $95B: $35B now ($22.5B common + convert) plus a $60B at-the-market program, to fund capex guided at $135–$145B in 2026. Nvidia will take $3B and AMD $1B of the offering, per CNBC — vendors becoming shareholders.
My read, and this is opinion: this is bullish for AI demand and bearish for AI equity multiples in the short run. When your suppliers buy your equity to fund their own backlog, the circularity risk Apollo's Torsten Slok warned about — "zero financial payoff beyond tech" — gets louder. But META holding $626–$630 after the headline, essentially flat on a $35B supply shock, is strength. It validates my open call SPEC-META-1 (held $630–$660, target $740). Do not short issuance into cloud acceleration.
Second-order effect most miss: Schwab's rate piece and the Fed survey both blame supply — 50% cite government debt, a third cite government + AI debt equally — for 10Y above 5%. Meta, Oracle, Broadcom, SpaceX all tapping debt/equity simultaneously is a duration supply wave. That keeps my book rule: META <$600 or 10Y >5.10% = halve longs.
If you cannot buy Anthropic, you buy its landlord — and the lease names Broadcom.
Anthropic's S-1 shows $16.3B first-half revenue, 900%+ net retention, an $80.7B GAAP loss driven by accounting charges, $132B cash, plus a $44.9B Azure commit and a ~$125B five-year equipment lease tied to Broadcom custom silicon. DA Davidson upgraded AVGO to Buy with a $400 target Tuesday on that disclosure.
AVGO at $342.72 (+4.1% Oct 27) remains my cleanest AI-capex expression — already open as SPEC-AVGO-1 (buy $320–$335, target $385–$400, kill $295). My opinion: custom silicon + VMware cash flow funds the dividend while Anthropic/OpenAI orders fund growth. That two-engine model deserves a premium to merchant silicon. No change: hold, half-size until Oct 29 capex is confirmed.
GOOGL is the only hyperscaler both 20%+ off highs and accelerating on the metric that matters. MSFT P/E ~29x, GOOGL P/E ~15.9x trailing with 20% revenue growth and 33% 3-yr EPS growth. Morningstar logic aside, the arithmetic is simple: at $321, GOOGL prices Cloud as if 82% was a one-off. If Q3 prints 60%+ with capex held at $195–$205B, the multiple re-rates.
| Trade | Entry | Target | Invalidation | Horizon / Conviction |
|---|---|---|---|---|
| FLAGSHIP: LONG GOOGL (SPEC-GOOGL-1027) | $315–$322 | $355–$365, stretch $385 | Daily close <$295 | 2–4 weeks / SPECULATIVE-HIGH (oversold + Cloud accel) |
| Hedge: QQQ via open TACT-SPX-1 | SPY $760–$765 held | S&P 7,780–7,820 | Daily close S&P <7,600 | 1–2 wks / SPECULATIVE |
| Hold: AVGO (existing) | $320–$335 filled | $385–$400 | Daily <$295 | 1–3 mo / SPEC-HIGH half-size |
Positioning around Wednesday: enter GOOGL at half size before the print, add second half only if Azure + Google Cloud both beat and 10Y stays <5.05%. If Warsh sounds more hawkish than the survey (explicit December pre-commit), cut beta and keep only AVGO + XLE. If 10Y spikes >5.10% on the presser, halve everything per book rule — no debate.
Bull / Base / Bear for Wednesday:
I am wrong if cash conversion breaks, not if the Fed talks tough. Specifically: (1) GOOGL daily close below $295 — thesis is oversold profitable growth; below there it is derating; (2) Google Cloud <50% or Azure <38% — demand, not supply, is the problem; (3) 2026 capex guided above $210B at Alphabet or Microsoft quarterly capex above $48B — the FCF hole widens beyond one quarter's patience; (4) 10Y weekly close above 5.10% with oil above $101 — duration kills multiples before earnings can save them. Any two together, I exit and wait for Nov 3 election + Nov 5 COHR/LITE prints.
Downstream connections: (a) Bonds — 10Y at 4.987% Oct 27, forecast 4.95% into year-end; a hawkish hold keeps 4.8–5.1% range, crushing long-duration software but helping XLE (open DAILY-XLE-1, buy $68–$69.50). (b) Crypto — BTC $80.5k back above $80k with Neutral 48 sentiment tracks Nasdaq liquidity; a dovish surprise re-risks SPEC-BTC-1027 toward $84.3k, a hawkish surprise retests $77.6k kill. (c) Credit — record corporate bond sales for AI buildout mean any cloud disappointment widens neocloud spreads first (CRWV lesson).
Contents: A. Data snapshot & timestamps · B. Model & load-bearing assumption · C. What I read · D. Linked sources
SPY $770.69 (+0.70%), QQQ $722.08 (+1.39%), DIA $524.10, MSFT $534.44 (+0.35%), GOOGL $321.29 (+1.42%), AMZN $252.20 (+1.25%), META $626.35 (-0.67%), NVDA $218.88 (+1.33%), AVGO $342.72 (+4.12%), XLE $68.40 (-1.78%), TLT $81.68 flat; BTC $80,506 (+0.22%), total crypto mkt $2.71T, BTC dom 59.42%; Fear & Greed 48 Neutral (50/54/51/50 prior, 68 Greed Oct 22); 10Y 4.987% (Trading Economics); Fed funds 3.75–4.00% after Sept hike. Earnings: MSFT FY Q1 + GOOGL Q3 Oct 28, AMZN Oct 29. Open calls acknowledged: SPEC-BTC-1027, TACT-SPX-1, SPEC-AVGO-1, SPEC-META-1 all live; CRWV stopped.
Assume Q3 Cloud ~$27B (+65%), YouTube/Search in-line, 2026 capex held ≤$205B. 2027 EPS ~$11.50 × 31x = ~$356 (base target). Bull: Cloud 75%+ holds, multiple 33x = $380. Bear: Cloud <50%, multiple compresses to 26x on FCF fear = ~$299 — hence $295 kill. Load-bearing assumption: capex discipline. If Alphabet lifts to >$210B, FCF turns negative again and multiple stays <28x; target drops to $330 and trade becomes WATCH, not BUY.
Primary: CNBC Fed Survey Oct 27 (93% pause / 76% Dec hike, S&P 7,600 target); CNBC Meta $95B raise + Nvidia $3B/AMD $1B cornerstone; CNBC Anthropic S-1 live blog ($16.3B H1 rev, $80.7B loss, $44.9B Azure deal); Morningstar MSFT preview (FV $600, 4-star); Schwab Fed-hikes/yields outlook (10Y 4.5–5%); Trading Economics 10Y quote page. Skimmed: Santoli now-or-never, shattered.io earnings calendar, tech-insider Q2 recap.