$GOOGL Cloud revenue +103%. Operating income +30%. Stock -10% after hours. Same night, Microsoft spent $45B on capex and rose ~4%. So the market is not punishing AI spending. My hypothesis: it is punishing who pays for it.
29 Oct 2026 · Alphabet Q3 2026 (unaudited release)
1) The number nobody led with
Q3 buybacks: zero. A year ago: $11.5B.
Year to date: zero, vs $39.8B in 2025.
Instead, Alphabet sold $12.4B of new stock in Q3 through an at-the-market program, and $42.9B this year.
Add $19.1B of convertible preferred and $85.2B of new debt: about $147B of outside money raised in nine months.
Diluted share count is up 1.35% year on year. The share count used to shrink every quarter.
2) Why it needed the money
Operating cash flow $42.3B. Capex $51.4B. Free cash flow -$9.1B, the second negative quarter in a row.
Capex guide for 2026 held at $195-205B. Year to date is $132B, so Q4 alone implies $63-73B (math is mine). And 2027 is guided to rise "significantly".
3) The staircase from "great quarter" to "miss"
- Revenue +25%, operating margin up from 30.5% to 31.8%. Fine.
- An $8.7B loss on equity holdings turned other income from +$12.8B to -$8.2B. That alone is roughly the gap between GAAP EPS $2.12 and adjusted $2.67 (math mine).
- Adjusted $2.67 still missed $3.02. Search +12% missed, blamed partly on AI Mode.
The operating business did its job. The per-share layer did not.
4) The Microsoft contrast
| Same night, same quarter | GOOGL | MSFT |
|---|---|---|
| Cloud growth | +103% (Google Cloud) | +48% (Azure) |
| Backlog / RPO | $561B | $677B |
| Quarterly capex | $51.4B | $45.2B (incl. leases) |
| Free cash flow | -$9.1B | positive, -27% YoY |
| Cash to/from shareholders | -$12.4B (stock sold), $0 buybacks | +$10.2B returned |
| Implied move | ~±6% | under ±5% |
| After-hours move | about -10.4% | about +4% |
Similar growth story. Opposite direction of cash between company and shareholder.
MSFT landed inside its implied move. GOOGL fell ~1.7x its implied move, the third report in a row it beat the options.
5) A worked example in made-up units
Company A: 100 shares, earns 10 units of cash, spends 12 on capex, issues 2 shares to cover it. Even if earnings grow 20% to 12, per-share earnings grow only 12/102 vs 10/100 = +17.6%. The 2.4-point leak is the cost of funding growth with owners' slices. Scale that to 2027 capex 'significantly' above $200B and the leak is the debate.
6) Method you can reuse on any hyperscaler
Operating cash flow − capex − (buybacks + dividends) = funding gap.
If the gap is filled by new shares, every backlog dollar is split among more owners. Growth per company ≠ growth per share.
Counter-case (I could be wrong)
Cloud margin went from 23.7% to 36.2%. Just over 55% of the $561B backlog lands within 24 months, about $154B a year vs a $123B run-rate today. Issuing stock near $320 to build assets earning ~36% margins can be smart capital allocation. If FCF turns positive in 2027 and the ATM stops, this sell-off was the wrong lesson.
My scenarios to the Q4 print (my view, not a forecast)
- 25%: FCF path improves, ATM paused, stock back above $322
- 50%: range $270-322 while 2027 capex is digested
- 25%: 2027 capex well above $300B with more issuance, below the $268 52-week low
IV usually falls ~29% within a week of GOOG reports, so defined-risk spreads fit this setup better than paying for raw volatility.
Dated tests
- Q4 call (early Feb 2027): 2027 capex number, and how much of the remaining $27.6B ATM was used.
- 10-Q: shares issued after 30 Sep.
Limits: release unaudited; I read call excerpts, not the full audio; the equity loss is not itemised by holding; price is after hours, not a cash close.
After reading this you can: check any AI spender's funding gap in three lines, and tell growth per company from growth per share.
Question: when a company sells stock to fund 36%-margin assets, is that dilution or a good trade for existing owners? Where would you draw the line?
Not financial advice. Scenarios and probabilities are my own view for education.
Sources: Alphabet Q3 2026 release (SEC 8-K Ex. 99.1), earnings call transcript (Investing.com), CNBC, Reuters, Microsoft FY27 Q1 release, Market Chameleon, Cboe LiveVol via CNBC. Links in the first comment.
#options #tradingpsychology #riskmanagement #optionstrading
Comments