Alphabet A (GOOGL)
The Cash Machine Invests More Than It Generates

On the evening of October 28, 2026, Alphabet reported numbers other companies would kill for: revenue +25%, Cloud more than doubled, operating income of $40.6B. The next day the stock lost 12.6% — its worst trading day since the 2004 IPO, roughly half a trillion dollars of market value gone in a single Thursday. That puts the stock almost 30% below its May record. This analysis replaces the August version and answers one question: what is noise here, and what is signal?

“The machine is running faster than ever. The market is asking a different question: how much of it still reaches the owner?”

1) Quick Overview

Alphabet is the holding company behind Google: Search, YouTube, Android, Chrome, Gmail and Maps, plus Google Cloud as the third-largest cloud provider and Waymo as a robotaxi operator in 15 U.S. cities. Just under 90% of all search queries worldwide run through Google (StatCounter, September 2026). Just under two thirds of revenue comes from advertising.

What happened on October 29

The drop has three causes. Only one of them is accounting.

• Earnings per share — mostly noise: $2.12 instead of the roughly $3 expected. Most of the gap comes from an $8.7B mark-to-market loss on equity stakes, according to Alphabet mainly the now publicly traded SpaceX stake. That is not cash leaving the company but a lower book value. It cut earnings by $0.55 per share. Without this effect EPS would have been $2.67 — still below the $3.02 analysts expected (LSEG).
• Free cash flow — signal: −$9.1B for the quarter, the second negative quarter in a row. The business generated $42.3B of operating cash flow, while capital spending came to $51.4B. For 2027, Alphabet expects capex to increase “significantly.”
• Search — signal with a question mark: Search revenue grew only 12%, after 17% in the prior quarter. Chief Business Officer Philipp Schindler names currency effects as the single largest driver, alongside retailers from Asia-Pacific, the end of the World Cup boost from the second quarter, and AI Mode, which so far brings in less advertising revenue per query. The market reads something else into it: the first measurable sign that AI is eating into the core business.

One quarter, two ledgers: earnings fall because of a book loss, free cash flow because spending exceeds cash flow. The first is noise, the second is signal. ↗ Click to enlarge

2) Business Model & Segments

The business model is a cycle, not a list. Billions of people use free products and pay with attention and data. That reach is sold to advertisers. The profit flows into AI infrastructure, which makes the products better — and builds new pillars on the side. The structure has not changed since August, but the proportions have.

The Alphabet machine as a cycle (fiscal 2025 figures): free products create reach, advertising creates cash, cash creates infrastructure — and that carries the new pillars. ↗ Click to enlarge

The third quarter of 2026 by revenue line ($B):

The shift is clear: a year ago Cloud stood for 15% of revenue, now it is 24%. Advertising is still growing, but only at low double digits. A large part of the group's growth now comes from Cloud.

New in the business model: TPU systems. Alphabet no longer only rents out its in-house AI chips (TPUs) inside its own cloud, it now sells complete systems of hardware, networking and software. TPU customers include Anthropic and Apple, and Google works with partners such as Oracle on delivery. The preparation shows up on the balance sheet: inventory rose from $2.4B to $13.3B since the start of the year, mostly hardware for TPU systems. The CFO expects a meaningful revenue ramp from this only in 2027.

The moat: not a single moat but a system of reinforcing advantages. Network effects (more users → better data → better results → more users), switching costs within the ecosystem (Android, Chrome, Gmail, Photos), economies of scale in data centers and in-house chips — and distribution power competitors would first have to replicate. Exactly this interlocking was the subject of two U.S. antitrust cases, see section 5.

3) Growth & Development

* Of which $8.06 are valuation effects on equity stakes (Q1 +$2.35, Q2 +$6.26, Q3 −$0.55, each after tax). Without them: $8.27 in nine months.

Revenue has grown in every single year, even during the 2022 rate shock, when earnings briefly dipped. In 2026 growth is even accelerating quarter by quarter: +22% in the first, +24% in the second, +25% in the third quarter. At 33.9%, the operating margin for the first nine months was above the prior year, despite billions in additional depreciation on new data centers. Quarter by quarter, however, it is falling: 36.1%, 34.0%, 31.8%.

The driver is Cloud: $30.7B of revenue in the third quarter (+103%), with a segment margin of 36.2% after 23.7% a year earlier. The backlog grew from $462B (March) to $514B (June) to $561B (September). Just over half of the total backlog is expected to turn into revenue within 24 months. Alphabet names one caveat itself: because its own capacity is tight, Google started renting third-party compute at the end of the third quarter — which will weigh on the Cloud margin in the near term.

Alien Analyzer V2 — Fair Value tab (Oct 30, 2026; German interface): after the drop the indicator reads “fairly valued (+5%)” — in August it was 34% above fair value. According to the tool the EPS line is estimated and includes the valuation gains, see section 6. ↗ Click to enlarge

4) Profitability & Balance Sheet

Alien Analyzer V2 — Quality check (Oct 30, 2026; German interface). Note: the Yahoo data partly still reflects June 30, 2026 for balance sheet and cash flow; the scorecard below uses the primary figures from the quarterly report. The automatic verdict “margin of safety” rests on the distorted P/E and does not apply here. The “buybacks” signal does not apply to 2026 either. ↗ Click to enlarge

On the data: the screenshot shows free cash flow of $22.7B, an equity ratio of 69.5% and 12.1 billion shares. These reflect older Yahoo data and, for free cash flow, a different definition. According to the quarterly report as of September 30, 2026, the figures are $19.7B of free cash flow over twelve months, a 67.5% equity ratio and 12.29 billion shares. As of October 30, 2026, the analyzer calculates free cash flow the way Alphabet does, as operating cash flow minus capital expenditures, and also arrives at $19.7B. The return on equity of 48.7% in the screenshot is also inflated by the valuation gains and is not a useful quality measure here.

The earnings side is strong: operating margin 31.8%, Cloud margin 36.2%. The cash side tells a different story. In 2025, $73.3B of free cash flow was left after all investments. In the first nine months of 2026 it was −$4.8B. Besides capital spending, the third quarter was also weighed down by the inventory build for TPU systems and the $5.2B payment for the EU antitrust fine from 2018.

The balance sheet: $265.1B of cash and marketable securities sounds like a fortress. But $90.4B of that are publicly traded shares, $75.2B in SpaceX stock alone, which cannot be sold in the short term. On the other side, long-term debt has risen from $46.5B to $125.1B since the start of the year. Excluding the equity stakes, a net cushion of roughly $49B remains after all financial debt [own calculation]. Equity ratio 67.5%, debt/equity 0.19: still a very solid balance sheet, but no longer a fortress balance sheet — and it is being rebuilt, from cash reserve to funding instrument.

On the dividend: in April, Alphabet raised its quarterly dividend to $0.22 (+5%), i.e. $0.88 per year and a 0.31% yield. It has not bought back a single share in 2026 — in 2025 buybacks were still $45.7B. The share count is rising again: from 12.09 to 12.29 billion since the start of the year.

5) Strategic Themes

The investment program: in February, Alphabet planned $175–185B of capital spending for 2026, in April $180–190B, in July $195–205B. The company confirmed that range in October. After $132.0B in the first nine months, the fourth quarter alone would need $63–73B. For 2027, CFO Anat Ashkenazi expects a “significant” increase, with details to follow alongside the full-year results in February. Market expectations of more than $300B are circulating (CNBC) [ESTIMATE — market expectation, not company guidance].

The funding: in June, Alphabet raised up to $84.75B of new equity. The building blocks: $18B of new common stock (half Class A at $355.20, half Class C at $351.80), $16.75B of mandatory convertible preferred stock with a 6.25% dividend and conversion in May 2029, a $10B private placement to Berkshire Hathaway, and an at-the-market program of up to $40B that lets Alphabet sell shares continuously on the exchange, mainly to cover tax obligations from employee stock compensation. That program brought in $12.4B net in the third quarter. On top come new bonds: long-term debt has risen 2.7-fold since the start of the year.

Equity stakes as a source of volatility: in the first half of 2026 Alphabet booked $135.9B of valuation gains on equity stakes, in the third quarter an $8.7B loss. The largest listed position is SpaceX (market value as of Sep 30: $75.2B). With Anthropic having filed for its IPO (Oct 26, 2026) — a company in which Alphabet reportedly holds around 14% — the next large book gain or loss may follow. For owners this means: GAAP earnings will now swing with other companies' share prices. Operating income is the more meaningful figure.

Antitrust — the risk has shrunk: in the ad tech case, Judge Brinkema ruled on September 2, 2026: no forced sale of the AdX ad exchange, instead behavioral remedies for six years under oversight. The final judgment is still pending. In the search case, the remedies without a Chrome sale stand (final judgment of December 5, 2025). Google has appealed, and so has the Justice Department, which wants to ban the permitted default-search payments to Apple and others. Oral argument will come in December 2026 at the earliest. In the EU, the 2018 Android fine has been final since the Court of Justice ruling of July 2, 2026 (paid: $5.2B including interest). In July, fines under the Digital Markets Act totaling €890M were added, and Google is challenging the 2025 EU ad tech fine of €2.95B before the EU General Court.

AI and Waymo: the Gemini app has had more than one billion monthly users since August. On the day of the results, Google introduced its new model Gemini 4 Carbon. Waymo now operates in 15 U.S. cities, provides more than 500,000 paid rides per week (as of September 2026, according to Bloomberg) and raised a $5B loan in October without an Alphabet guarantee.

Outlook (analyst consensus, stockanalysis.com, as of Oct 29, 2026): revenue 2026E $499B (+24%), 2027E $634B (+27%). EPS 2027E $15.17 [ESTIMATE — analyst forecast, adjusted according to the source; whether valuation effects are included is not disclosed]. I deliberately leave out the 2026 consensus: at $16.26 it is below what has already been booked in nine months, so it apparently mixes booked valuation gains with adjusted figures.

6) Valuation in Context

Alien Analyzer V2 — Multiples tab (Oct 30, 2026; German interface: KGV = P/E, KBV = P/B, KUV = P/S, KCV = P/CF): P/E 14.8 below the 10-year average of 19.6 (Fair Value and Quality check tabs: 20.1) — and the warning “many P/E outliers filtered.” Using the figures as of Sep 30, P/B and P/S are slightly lower (5.2 and 7.4). ↗ Click to enlarge

Here lies the most important lesson, just as in August. A trailing P/E of 14.8 looks like a fire sale — a world-class company well below its own ten-year average. But it is an optical illusion: roughly $8 of the $19.15 of earnings per share over the last twelve months are valuation gains on equity stakes, not income from the business. They inflate the earnings in the denominator of the P/E.

Three angles on the same price of $283.82 [ESTIMATE — own calculations, not facts]:

• Trailing, as reported: P/E 14.8 — flattered by one-off effects.
• Operating, without valuation effects: $8.27 in nine months, annualized about $11 — P/E around 26. That is above the stock's own ten-year average.
• Forward, 2027 consensus: $15.17 per share — P/E around 18.7. This assumes earnings per share rise by a good third above the 2026 operating level in 2027, despite higher depreciation and more shares.

In August the stock stood at $340.67, about 23 times the 2027 consensus at that time. Today it is 18.7 times. A large part of the premium the market paid for Alphabet over the summer is gone. The analyzer's fair value indicator shows the same: “fairly valued (+5%)” instead of “34% above fair value.” That does not make it cheap in the sense of a value stock — on an operating basis it still costs more than its own average, and at 7.4 the P/S ratio is more than twice the ten-year average. Whether that is justified hinges on a single question: will the investments earn their return?

7) Competitive Landscape & Moat

Sources: StatCounter (search, September 2026), Synergy Research (cloud, Q3 2026), Google and OpenAI (user numbers). Weekly and monthly user numbers are not directly comparable.

The front line is not Bing, it is behavior. Google's share of search queries remains just under 90%. The question is how much a query is still worth when users get answers instead of links. ChatGPT reaches 1.2 billion weekly users, Gemini more than one billion monthly. Alphabet defends its monopoly by rebuilding its own business model: AI answers directly in search, AI Mode as a separate mode. The third quarter shows the price of that. Usage is growing, but AI Mode so far brings in less advertising revenue per query. Whether that is a transition phase or a permanently lower level is the central open question for the stock.

In the cloud, Google is catching up: according to Synergy Research, AWS held a 26% market share in the third quarter of 2026, Microsoft 19%, Google 16% and rising. The structural advantage is the in-house TPU chips. Alphabet builds AI models, chips and data centers from a single source — and now even sells those chips to others.

8) Customer Perspective

Alphabet's paying customers are not the billions of users but advertisers and cloud customers. For advertisers, Google Search is the place where purchase intent arises. This lack of alternatives is the core of its pricing power — and was at the heart of the U.S. search case.

In the cloud business, the $561B backlog and large customers such as Anthropic and Apple speak for themselves. Public B2B satisfaction scores in the classic sense do not exist for either customer group. [NOT AVAILABLE]

9) Employee Perspective

At the end of September 2026 Alphabet employed 204,757 people, about 14,600 more than a year earlier. For two decades the company has been one of the most sought-after employers in tech, with high pay and generous stock packages — so generous that part of the share sale program funds the resulting tax bill. In the AI era, top researchers are the industry's most contested resource. Even Alphabet loses prominent minds: chief scientist Jeff Dean left in August 2026 after 27 years. Systematic employer rating data could not be evaluated for this analysis. [NOT AVAILABLE]

10) Opportunities & Risks

11) Alien Verdict

Alien Analyzer V2 — Dividend tab: the history only begins in 2024. The dividend is growing, buybacks have been on hold since the start of the year. ↗ Click to enlarge

Operationally, the third quarter of 2026 was one of the strongest in Alphabet's history: 25% revenue growth on $127.6B of quarterly revenue, a cloud business that doubles while becoming more profitable, operating income of $40.6B. The drop in earnings per share is mostly accounting.

What remains is more serious: Alphabet has turned from a cash distributor into a capital consumer. No more buybacks, a rising share count, debt up 2.7-fold, two quarters of negative free cash flow, and capital spending set to rise significantly again in 2027. At the same time, Search — the business that pays for all of this — is growing more slowly. That is no proof that AI is destroying the search business. But it is the first data point pointing in that direction, and Alphabet names currency effects as the largest driver — an explanation the next quarter has to confirm or refute.

The share price has absorbed part of this. Measured against the 2027 consensus, the stock is noticeably cheaper today than in August; measured against today's operating earnings, it is not cheap. Anyone holding Alphabet is betting that the largest investment program in the company's history will earn its return — through Cloud, TPU sales and an AI search that ultimately makes money as well as the old one. The next checkpoints are concrete: the 2027 capex figure in February, search growth excluding currency effects, and the point at which free cash flow turns positive again.

“A company that invests more than it earns is not automatically in danger. It is a bet. The only question is whether the owner knows he is placing it right now.”

Data and sources: prices up to the close of Oct 29, 2026 (Nasdaq). Analyzer screenshots from Oct 30, 2026 (Yahoo Finance data). Quarterly figures from Alphabet's third-quarter 2026 earnings release (SEC Form 8-K, Oct 28, 2026), the 10-Q quarterly report (Oct 29, 2026) and the earnings call. Antitrust status per the 10-Q, court announcements and the European Commission. Consensus: stockanalysis.com, Oct 29, 2026. Values marked [ESTIMATE] are own calculations or third-party forecasts, not facts.