Alphabet
This is the archived full research. Read the story
The company behind Google Search, YouTube and Android, paid for by advertisers, which also rents computing power and AI to other businesses through Google Cloud.
The scores
Growth4/4 ↑
Revenue up 25% and operating income up 30%, with Cloud up 103% on a broadening base; search slowed to 12% from 17%.
Business quality4/4 →
About 90% of worldwide search held through three years of chatbot attack, against Bing near 5%, and search revenue accelerated; distribution is now court-restricted.
Financial position3/4 ↓
$265.1B of cash and securities against $125.1B of debt, but free cash flow is negative and $707.0B of commitments sit off the balance sheet.
Execution record3/4 →
Cloud margin held at 36% and Gemini 4 shipped, but the run of earnings beats ended at eight and the spending guide rose twice.
Per share capture2/4 ↓
Buybacks stopped, 86M new shares in June and 36M more in Q3, a convertible landing in 2029; the count keeps rising after years of falling.
Potential3/4 ↑
15% of a $500B cloud market growing 43%, with $513.9B already signed; the upside needs the AI build to earn a return nobody has shown.
Risk profile2/4 →
Appeals in the search case and EU orders can each reshape distribution in a ruling, and earnings swing with a $94.1B SpaceX stake.
Valuation3/4 ↑
About 20.7 times operating profit with stakes netted, and about 93% of a three-stage value near $307: the price pays for the build working.
Evidence quality3/4 →
Revenue, profit, backlog and financing are filed; share and market size come from outside measurers; AI usage rests on management alone.
The Verdict
We rate Alphabet Neutral. Google Search still handles about 93% of the world's searches and Google Cloud is booming, so the business is strong. But Alphabet is spending about $200B this year on AI data centres, its cash flow has turned negative and it has started selling new shares. Valued on its own future cash, with growth slowing to 4% and a margin that pays for the build, the shares are worth about $307 against $284: fair, not cheap. The view hinges on whether the build earns its return in cash.
The long caseCompanies and AI labs want more computing than providers can build, and AI is changing search; Alphabet's huge data-centre spending must earn its return in cash.
How we weighed it
The price against our value
Cuts both ways
Price at 93% of our value
The price is inside the margin of safety on its future cash, with the bear case well below.
The business
In its favour
Solid
Search still dominates and the cloud is booming.
Growth against what the price assumes
In its favour
Search and cloud growth support what the price assumes.
What could go wrong
Against it
Chatbots may take searches, the spending may not pay, courts may reshape distribution and its models have slipped.
Against competitors
Cuts both ways
Just below the middle of its rivals, in line with the group.
How its group is behaving
Cuts both ways
The largest technology names move together on AI spending; its building bill is its own.
The latest news and report
Cuts both ways
Spending keeps rising and cash stays under pressure, while cloud demand held up.
The evidence under this - 9 claims
- 2026-10-28editedQ3 2026 release replaces the Q2 revenue lines
- 2026-10-28editedQ3 2026 cash flow statement and free cash flow reconciliation
Management says the raise reflects faster delivery of capacity to meet demand already signed, and the backlog rose more than $50B in the same quarter (C09).
- 2026-10-29editedQ3 2026 call: the guide held, 2027 up significantly
- 2026-10-28editedQ3 2026 release: the share program was used, $12.4B in one quarter
- 2026-10-29editedQ3 2026 10-Q wording of the programme's use
StatCounter counts page views from search engines only, so a question asked of ChatGPT instead of Google does not appear in its denominator at all.
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
- 2026-11-09editedrestated on enterprise value net of securities and investments, with preferred and minority interest added; the vendor's figure subtracts cash alone
- 2026-11-09editedRe-struck on the decaying fade the owner chose on 9 November: after the stated path each year closes 30% of the gap to 4%, and an analysts' margin is cut by their gap to the filed operating profit.
The P/E makes Alphabet the cheapest of the four, but only because the $98.0B equity gain (C06) inflates its earnings; on operating profit it is second cheapest, still dearer than Meta.
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
- 2026-11-09editedrestated on enterprise value net of securities and investments, with preferred and minority interest added; the vendor's figure subtracts cash alone
Operating profit grew 30% in Q3 2026 (C04), but depreciation on about $200B a year of building (C11) has yet to land on it, so today's profit flatters what the build will earn.
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
Whether about $200B a year of building earns its return cannot be seen from outside (C11, C12): at 10% and 6% growth after 2026 and a 26% margin the value is about $162. The 30% margin assumes depreciation does not keep climbing past 2028.
- 2026-11-09editedRe-struck on the upgraded method: every view read at 9%, the return on new capital set by the moat, and capital per dollar of sales read from the filings.
- 2026-11-09editedRe-struck at the 8% required return the owner chose after the roll-out screen: the market priced the median name on the roster at about 7.7%, so 9% on top of a 20% margin read most of it Unattractive.
- 2026-11-09editedRe-struck on the decaying fade the owner chose on 9 November: after the stated path each year closes 30% of the gap to 4%, and an analysts' margin is cut by their gap to the filed operating profit.
- 2026-11-10editedre-valued on the owner's rules of 10 November: the return on new capital glides to its terminal level over years six to ten
The company
Alphabet gives away Google Search, YouTube, Android, Chrome and Gmail to billions of people. Advertisers pay to appear beside what those people look for and watch, bidding in an auction each time a search is made.
Its second business, Google Cloud, rents data-centre computing, Google's own AI chips and its Gemini AI models to companies. In Q2 2026 it began selling whole racks of those chips for customers' own buildings. Without it, advertisers would turn to Meta, TikTok or Amazon, and cloud buyers to Amazon or Microsoft.
The evidence under this - 4 claims
- 2026-10-28editedQ3 2026 release replaces the Q2 revenue lines
How it makes money
Most dollars arrive one click at a time: an advertiser pays Google when someone clicks an ad beside a search. In Q2 2026 clicks rose 13% and the price of each click 3%. Cloud customers pay for computing as they use it, often under multi-year contracts. In Q3 2026 Google Services kept 40.7 cents of each sales dollar as operating profit; Cloud kept 36.2 cents.
The evidence under this - 4 claims
- 2026-10-28editedQ3 2026 release replaces the Q2 revenue lines
- 2026-10-28editedQ3 2026 segment table replaces Q2
Contracts of one year or less were added to the backlog definition from Q1 2026, so part of the growth is a change in what is counted.
- 2026-10-29editedQ3 2026 10-Q, filed after the first read: backlog $567.7B
Where the money comes from, stream by stream
| Stream | Share of sales (Q2 2026) | How it earns it |
|---|---|---|
| Google Search & other | 52.8% | Search ads sold by auction on Google Search, Maps, Gmail and Play; advertisers pay per click. |
| Google Cloud | 20.7% | Rented computing, AI models and TPU systems, data tools and Workspace subscriptions sold to businesses and governments. |
| Google subscriptions, platforms and devices | 10.8% | YouTube and Google One subscriptions, Play store fees on apps and Pixel and other hardware. |
| YouTube ads | 9.2% | Video and Shorts ads sold to advertisers, priced by impression and view. |
| Google Network | 6.1% | Ads placed on other publishers' sites and apps through AdSense, AdMob and Ad Manager; Google keeps a share. |
| Other Bets and hedging | 0.4% | Waymo rides, GFiber internet and Verily health services, plus foreign-exchange hedging gains. |
Who runs it
Sundar Pichai has run Google since 2015 and its parent Alphabet since 2019; Anat Ashkenazi, from Eli Lilly, has run its finances since 2024. The co-founders Larry Page and Sergey Brin still control most of the votes through super-voting shares, so outside shareholders cannot outvote them. Top executives get no cash bonus: most of their pay is stock, and the part that depends on results pays only on how the shares do against other large companies, not on profit or cash.
The evidence under this - 3 claims
- No element pays on revenue, margin, free cash flow or return on capital, so the capital spending build in C12 and the share sales in C13 cost the plan nothing directly; only the share price does.
- The proxy asked shareholders for 200M more Class C shares for the stock plan, so the dilution in C16 is planned to continue.
Why it wins, and who could beat it
Why customers come back
Google owns the doors people search through. Android and Chrome are its own, and it pays to be the default search on other browsers and phones. Every search teaches it to rank results and price ads better, which pays for more defaults.
It has held up. StatCounter measured Google at 89.94% of worldwide searches in September 2026, barely changed in a year in which every big chatbot launched. A US court has now limited how Google buys those default spots, which is the mechanism to watch.
The evidence under this - 4 claims
A default captures the user who opens a browser; a user who opens a chatbot app instead never meets it (C20), so distribution protects the old door, not a new one.
StatCounter counts page views from search engines only, so a question asked of ChatGPT instead of Google does not appear in its denominator at all.
The threat
The threat is that people ask a chatbot instead of searching. Apple said in 2025 that Safari searches fell for the first time, and it added OpenAI, Perplexity and Anthropic as options. Advertisers' money is moving too: in the US, social media ads grew 32.6% in 2025 against 11% for search. In Cloud, Amazon and Microsoft are larger.
The evidence under this - 4 claims
Google answered the same day that its total queries, including those from Apple devices, were still growing (S10).
Google's own search revenue grew faster than the US search market, 17% in Q2 2026 (C02), so it may be taking share within a slower pool.
Google owns YouTube, whose ads sit in IAB's video line, so part of the money leaving search is still Google's (C02).
The rivals: META, MSFT, MSFT, AMZN, ORCL, CRWV, and 3 more that are private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | Where it competes | How far along | Priced against it |
|---|---|---|---|---|
| Private | OpenAI's ChatGPTPriced against it: Private | Google Services | On sale | Private |
| Private | Perplexity and Anthropic's ClaudePriced against it: Private | Google Services | On sale | Private |
| META | Meta's Facebook and Instagram adsPriced against it: Yes | Google Services | On sale | Yes |
| Private | TikTokPriced against it: Private | Google Services | On sale | Private |
| MSFT | Microsoft Bing and CopilotPriced against it: Yes | Google Services | On sale | Yes |
| Microsoft Azure | Google Cloud | On sale | ||
| AMZN | Amazon Web ServicesPriced against it: Yes | Google Cloud | On sale | Yes |
| ORCL · CRWV | Oracle and CoreWeavePriced against it: One segment only | Google Cloud | On sale | One segment only |
How big it could get
How big this could get
Renting computing over the internet is a $500B-a-year market that grew 43% in Q2 2026, according to Synergy Research. Google holds 15% of it, behind Amazon and Microsoft. It has $513.9B of Cloud contracts already signed, about the size of the whole market's yearly sales. Nobody publishes chips installed or prices paid, so the market can only be bounded.
The evidence under this - 2 claims
Contracts of one year or less were added to the backlog definition from Q1 2026, so part of the growth is a change in what is counted.
- 2026-10-29editedQ3 2026 10-Q, filed after the first read: backlog $567.7B
Who has signed
Cloud customers have signed $513.9B of contracts, just over half due within two years. A few buyers have also signed for Google's chip racks, mostly paid in 2027, with no size given. The filings name none of these customers, so how much rests on a few AI labs cannot be checked. Advertising, three-quarters of sales, has no contracts at all.
The evidence under this - 3 claims
Contracts of one year or less were added to the backlog definition from Q1 2026, so part of the growth is a change in what is counted.
- 2026-10-29editedQ3 2026 10-Q, filed after the first read: backlog $567.7B
The 10-Q does not split the backlog by customer, so how much rests on a few AI labs cannot be checked from the filing.
Where the growth is coming from, business by business
Where the growth is coming from
| Business | Share of revenue | Growth: last quarter | Last half-year | Last year | Share in 3 years at its half-year rate |
|---|---|---|---|---|---|
| Google Services | 75% | +10.6% | +12.5% | +13.7% | 37% |
| Google Cloud | 24% | +102.6% | +92.8% | +75.8% | 60% |
| Rest of the company | 0% | +313.1% | +164.8% | +3.9% | 3% |
| Whole company | 100% | +24.7% | +24.4% | +22.2% | 100% |
On the filed split through 2026-09-30, if each business keeps its last half-year's growth, the company grows about 33% next year and 53% in year 3, against 22.2% over the last twelve months. Google Cloud grows fastest: it grows 93% against 13% for the rest, and would be 60% of revenue by then. A scenario, not a forecast: no business holds its rate.
How we read it: The business carrying the case: Google Cloud grows about eight times faster than the rest on a $513.9B backlog. The 4 does not need it; search alone grew 12%.
The results
The numbers
Sales grew 22% over the twelve months to Sep 2026, faster than the 15% in the year to Dec 2025.
Operating margin is 33%, up from 27% in the year to Dec 2023.
Running costs take 28¢ of each sales dollar, up from 26¢ in the year to Dec 2021.
Each dollar put into the business earns 21% a year after tax, about where it was (26%) in the year to Dec 2021.
It turned 4.2% of sales into free cash, $19.7B over the last twelve months, after spending 34% of sales on buildings and equipment.
It holds $265B of cash against $126B of debt, $139B more cash than it owes.
The share count fell 6.0% in 3.8 years, and pay in shares costs 7.0% of sales.
All the numbers
Filed figures from the company's 10-K and 10-Q, read 2026-11-10. Fiscal years end in December.
| Dec 2021 | Dec 2022 | Dec 2023 | Dec 2024 | Dec 2025 | 12 mo to Sep 2026 | |
|---|---|---|---|---|---|---|
| Size and growth | ||||||
| Sales | $258B | $283B | $307B | $350B | $403B | $471B |
| Sales growth | +41% | +9.8% | +8.7% | +14% | +15% | +22% |
| Profit | ||||||
| Gross margin | 57% | 55% | 57% | 58% | 60% | 61% |
| Operating margin | 31% | 26% | 27% | 32% | 32% | 33% |
| Net margin | 30% | 21% | 24% | 29% | 33% | 50% |
| What growth costs, per sales dollar | ||||||
| Research and development | 12% | 14% | 15% | 14% | 15% | 16% |
| Sales and marketing | 8.9% | 9.4% | 9.1% | 7.9% | 7.1% | 7.0% |
| General and admin | 5.2% | 5.6% | 5.3% | 4.1% | 5.3% | 4.7% |
| Pay in shares | 6.1% | 6.9% | 7.2% | 6.5% | 6.7% | 7.0% |
| Capital spending | 9.6% | 11% | 10% | 15% | 23% | 34% |
| Cash | ||||||
| Cash from operations | $91.7B | $91.5B | $102B | $125B | $165B | $180B |
| Free cash flow | $67.0B | $60.0B | $69.5B | $72.8B | $73.3B | $19.7B |
| Balance sheet | ||||||
| Cash and short-term investments | $140B | $114B | $111B | $95.7B | $127B | $265B |
| Debt | $14.8B | $14.7B | $12.9B | $11.9B | $48.5B | $126B |
| Net cash | $125B | $99.1B | $98.0B | $83.8B | $78.3B | $139B |
| Returns and shares | ||||||
| Return on capital | 26% | 23% | 26% | 30% | 27% | 21% |
| Shares (diluted) | - | 13.2B | 12.7B | 12.4B | 12.2B | 12.4B |
What growth costs
Growth has made Alphabet more profitable: selling costs and the cost of serving each sale shrank as revenue grew, while research spending rose only a little faster than sales. The cost growing fastest is building data centres, which has nearly trebled for each dollar of sales and is only starting to show up as a cost in profit.
The evidence under this - 2 claims
- Capital spending is not an expense when paid: depreciation rose from $9.5B to $13.6B in a half year (C43), so the trebled capital spending per sales dollar has yet to reach the margin.
- The R&D line in the numbers card is the filer's ResearchAndDevelopmentExpense tag and matches the income statement's Research and development line ($61,087M in 2025).
The last report (28 Oct)
Quarter ended September 2026, filed figures.
Sales rose 25% to $127.6B and operating profit 30%. Google Cloud grew 103%, with 36.2 cents of each sales dollar kept as profit, and search grew 12%, down from 17%. YouTube ads grew only 9%.
Reported earnings of $2.12 a share fell from $2.87, because investments lost $8.7B in value. The cash side worsened: $51.4B went on data centres in the quarter, free cash flow was a $9.1B outflow, debt reached $125.1B, and Alphabet sold another $12.4B of shares.
On the call, management left 2026 spending unchanged, said 2027 will rise significantly, and blamed slower search mostly on currency, weaker retail and media spending and cheaper-earning AI Mode searches.
What it did to the thesisThe call did not change the view. Spending keeps rising and cash stays under pressure, while Cloud demand and the search explanation held up.
The quarter in detail
Sales growth from the previous quarter, filed figures.
Earnings per share $2.12 against $3.01 expected; $2.12 as filed, with no adjustments.
The evidence under this - 7 claims
- 2026-10-28editedQ3 2026 release replaces the Q2 revenue lines
- 2026-10-28editedQ3 2026 segment table replaces Q2
- 2026-10-28editedQ3 2026 cash flow statement and free cash flow reconciliation
- 2026-10-28editedQ3 2026 release: the share program was used, $12.4B in one quarter
- 2026-10-29editedQ3 2026 10-Q wording of the programme's use
- 2026-10-28editedQ3 2026 balance sheet and release
The inventory build and the $5.2B EC fine pushed Q3 free cash flow to negative $9.1B; management gave no ex-TPU Cloud growth rate and no TPU margin.
Contracts of one year or less were added to the backlog definition from Q1 2026, so part of the growth is a change in what is counted.
- 2026-10-29editedQ3 2026 10-Q, filed after the first read: backlog $567.7B
What management did with the money
Alphabet bought back $45.7B of its shares in 2025 and none in 2026. It sold 86M new shares in June and 36M more in Q3, issued preferred shares that turn into common stock in 2029, and borrowed heavily: debt rose to $125.1B from about $16B a year earlier.
Insiders in 2026Q3: 79 purchases and 106 sales, from their Form 4 filings.
The evidence under this - 3 claims
- 2026-10-28editedQ3 2026 release: still no buyback, count higher
- 2026-10-28editedQ3 2026 release: the share program was used, $12.4B in one quarter
- 2026-10-29editedQ3 2026 10-Q wording of the programme's use
- 2026-10-28editedQ3 2026 balance sheet and release
The price, and where we disagree
In short. Fair. The price is about what the business is worth if its AI spending pays off roughly as planned, with little paid for anything beyond it.
The Q3 2026 results on 28 October showed Cloud doubling but free cash flow below zero, and the shares fell sharply. By 2 November they were 18% below their October price, at about 22 times operating profit.
In 2025 the market feared chatbots would kill search, and was wrong. Now it is weighing whether $200B a year of building pays back.
The evidence under this - 7 claims
- 2026-10-28editedQ3 2026 release replaces the Q2 revenue lines
- 2026-10-28editedQ3 2026 cash flow statement and free cash flow reconciliation
Management says the raise reflects faster delivery of capacity to meet demand already signed, and the backlog rose more than $50B in the same quarter (C09).
- 2026-10-29editedQ3 2026 call: the guide held, 2027 up significantly
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
- 2026-11-09editedrestated on enterprise value net of securities and investments, with preferred and minority interest added; the vendor's figure subtracts cash alone
- 2026-11-09editedRe-struck on the decaying fade the owner chose on 9 November: after the stated path each year closes 30% of the gap to 4%, and an analysts' margin is cut by their gap to the filed operating profit.
The P/E makes Alphabet the cheapest of the four, but only because the $98.0B equity gain (C06) inflates its earnings; on operating profit it is second cheapest, still dearer than Meta.
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
- 2026-11-09editedrestated on enterprise value net of securities and investments, with preferred and minority interest added; the vendor's figure subtracts cash alone
How we value it
- Method
- Cash the business will produce, in three stages
- Required return
- 8%
- Long-run growth
- 4%
- Moat
- A strong moat
| Bear | Base | Bull | |
|---|---|---|---|
| Sales growth, years 1 to 3 | 24%, 10%, 6% | 24%, 20%, 16% | 24%, 27%, 22% |
| Normal operating margin | 26% | 30% | 34% |
| Return on new capital after year ten | 8% | 15% | 15% |
| Value a share | $162 | $307 | $410 |
- Price
- $284
- Priced on
- 6 Nov
- Price as a share of the base value
- 93%
- Starting growth the price pays for
- 23.2%
- Yearly return the price offers on the base case
- 8.3%
Where each figure comes from
| Figure | Source | Where exactly |
|---|---|---|
| Sales, the last full year | Filed report | Annual report (10-K) for the period to 31 Dec 2025, filed 5 Feb 2026; the full year. |
| Operating margin today | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the four quarters to that date. |
| Tax rate | Our judgement | Our reading of the evidence. |
| Sales per dollar of new capital | Filed report | Annual report (10-K) for the period to 31 Dec 2025, filed 5 Feb 2026: the year's sales over its operating capital. |
| Shares | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the diluted share count. |
| Cash and investments | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the balance sheet on that date. |
| Debt | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the balance sheet on that date. |
| Sales growth, years 1 to 3 | Our judgement | Our reading of the evidence; each case below names it. FMP's consensus of 17 to 29 analysts read 27.3%, 23.7%, 17.4%. |
| Normal operating margin | Our judgement | Our reading of the evidence; each case below names it. FMP's consensus, less stock pay, read 26.9%. |
| Return on new capital after year ten | Our method's rule | Set by our reading of the moat: the 8% rate with none, 12% with an average one, 15% with a strong one; the bear case earns the rate. |
| The bear and bull cases | Our judgement | Our reading of the evidence; each case below names it. |
| Price | Market price | FMP's closing price on 6 Nov 2026. |
| Preferred stock | Filed report | Read by hand from the filed statements; the period it was read from was not recorded. |
| Other investments | Filed report | Read by hand from the filed statements; the period it was read from was not recorded. |
What each case rests on
Bear caseThe build runs ahead of demand: after 2026 growth falls to 10% and 6% as search slows below 10% and Cloud customers stop exceeding their commitments (C09), the margin settles at 26% under the depreciation, and new capital earns only the 9% it costs.
Base case2026 grows about 24% to near $500B, the year to September having grown 25% with Cloud up 103% (C02); then 20% and 16% as about 55% of the $560.8B Cloud backlog lands within two years (C08); the operating margin settles at 30%, below today's 33.3%, once depreciation on the $195-205B build lands (C12).
Bull caseCloud keeps growing above 60% while TPU system sales ramp in 2027 (C50), lifting growth to 27% and 22%, and scale holds the margin near 34%, about where the year to September 2026 ran (C05).
What moves the base value
| Lower | Higher | |
|---|---|---|
| Growth in years 1 to 3, a quarter lower or higher | $261 | $359 |
| The normal margin, 15% lower or higher | $264 | $349 |
| Capital per dollar of growth, a quarter more or less | $300 | $311 |
| Return on new capital after year ten, at the rate or one step up | $231 | $307 |
| The rate, a point higher or lower | $247 | $406 |
| Long-run growth a point lower | $272 | - |
Value a share as growth and margin change
| Normal margin ↓ · growth → | 50% | 75% | 100% | 125% | 150% |
|---|---|---|---|---|---|
| 24% | $184 | $215 | $250 | $291 | $339 |
| 27% | $204 | $238 | $278 | $325 | $380 |
| 30% | $223 | $261 | $307 | $359 | $421 |
| 33% | $242 | $285 | $335 | $393 | $461 |
| 36% | $261 | $308 | $363 | $428 | $502 |
Growth is shown as a share of the base case's. Green: the price is at or under 80% of that value. Red: at or over 120%.
How we read the multiple
Buyers pay about 21 times last year's operating profit for the whole company, a little below its rivals' middle and above Meta. That profit leaves out the SpaceX gain, so it is the figure to trust.
- After tax, the price is about 26 times what the business earned from operations.
- Over six years, buyers usually paid about 25 times earnings.
- Valued with growth slowing to 4% and a 30% margin that pays for the build, the shares are worth about $307.
Against its own history. Close to its own past on what the business itself earns. The cheap-looking earnings figure is inflated by investment gains, so the stricter reading is the one to trust.
The evidence under this - 8 claims
- 2026-10-28editedQ3 2026 segment table replaces Q2
- 2026-10-28editeda fourth quarter rolled into the trailing year
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
- 2026-11-09editedrestated on enterprise value net of securities and investments, with preferred and minority interest added; the vendor's figure subtracts cash alone
- 2026-11-09editedRe-struck on the decaying fade the owner chose on 9 November: after the stated path each year closes 30% of the gap to 4%, and an analysts' margin is cut by their gap to the filed operating profit.
The P/E makes Alphabet the cheapest of the four, but only because the $98.0B equity gain (C06) inflates its earnings; on operating profit it is second cheapest, still dearer than Meta.
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
- 2026-11-09editedrestated on enterprise value net of securities and investments, with preferred and minority interest added; the vendor's figure subtracts cash alone
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
Operating profit grew 30% in Q3 2026 (C04), but depreciation on about $200B a year of building (C11) has yet to land on it, so today's profit flatters what the build will earn.
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
Whether about $200B a year of building earns its return cannot be seen from outside (C11, C12): at 10% and 6% growth after 2026 and a 26% margin the value is about $162. The 30% margin assumes depreciation does not keep climbing past 2028.
- 2026-11-09editedRe-struck on the upgraded method: every view read at 9%, the return on new capital set by the moat, and capital per dollar of sales read from the filings.
- 2026-11-09editedRe-struck at the 8% required return the owner chose after the roll-out screen: the market priced the median name on the roster at about 7.7%, so 9% on top of a 20% margin read most of it Unattractive.
- 2026-11-09editedRe-struck on the decaying fade the owner chose on 9 November: after the stated path each year closes 30% of the gap to 4%, and an analysts' margin is cut by their gap to the filed operating profit.
- 2026-11-10editedre-valued on the owner's rules of 10 November: the return on new capital glides to its terminal level over years six to ten
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on company value for each $1 of operating profit, cheapest first:
| Ticker | Company | company value for each $1 of operating profit |
|---|---|---|
| META | Meta Platforms | 18.6× |
| GOOG | Alphabet | 22.2× |
| MSFT | Microsoft | 24.4× |
| AMZN | Amazon.com | 28.1× |
The middle of the other 3 is 24.4. Alphabet Inc. is 2nd cheapest of 4.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
Just below the middle of its rivals and dearer than Meta, so the price is in line with the group; the whole group has cheapened since October.
Where we disagree with the market
Alphabet's AI spending is running ahead of its cash: the shares fell 18% in a month after the Q3 results showed free cash flow below zero.
The business beat the 2025 fears, and after the fall the price pays for the AI build working as planned, though Alphabet now sells its shares instead of buying.
Twice the market sold Alphabet on threats to search, in early 2025 over chatbots and later that year over a forced sale of Chrome, and both times it was wrong. This time it is selling on the cost of the build.
The case against us is cash: free cash flow fell below zero in Q3 2026 and spending rises again in 2027. If the build does not pay, a fair price on today's profit is not a cheap one.
What analysts say, and the evidence
The evidence under this - 9 claims
- 2026-10-28editedQ3 2026 segment table replaces Q2
- 2026-10-28editedQ3 2026 cash flow statement and free cash flow reconciliation
Management says the raise reflects faster delivery of capacity to meet demand already signed, and the backlog rose more than $50B in the same quarter (C09).
- 2026-10-29editedQ3 2026 call: the guide held, 2027 up significantly
- 2026-10-28editedQ3 2026 release: the share program was used, $12.4B in one quarter
- 2026-10-29editedQ3 2026 10-Q wording of the programme's use
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
- 2026-11-09editedrestated on enterprise value net of securities and investments, with preferred and minority interest added; the vendor's figure subtracts cash alone
- 2026-11-09editedRe-struck on the decaying fade the owner chose on 9 November: after the stated path each year closes 30% of the gap to 4%, and an analysts' margin is cut by their gap to the filed operating profit.
Operating profit grew 30% in Q3 2026 (C04), but depreciation on about $200B a year of building (C11) has yet to land on it, so today's profit flatters what the build will earn.
- 2026-11-03editedre-struck at the 2 November close of $278.04 on Q3 2026 trailing profit; route P1 fired
What could go wrong, and how we would know
The risks
- Chatbots take the searches. If people ask ChatGPT or Claude instead of Google for things they buy, the clicks that pay for everything fall, and StatCounter would not show it.
- The spending does not pay. About $200B this year and more in 2027 is committed before the revenue that would repay it, with $707.0B of supply commitments behind it.
- Courts reshape distribution. Both sides have appealed the search ruling, and European orders could open Android and search data to rivals.
- Falling behind on AI models. Gemini 3.5 Pro was abandoned and Gemini 4 reached only trusted partners, after reports that Google trailed OpenAI and Anthropic in coding.
- More new shares. Alphabet sold $12.4B of shares in Q3 under a $40.0B programme and has $27.6B left, and preferred shares convert in 2029, so the count can keep rising.
The evidence under this - 8 claims
Google answered the same day that its total queries, including those from Apple devices, were still growing (S10).
StatCounter counts page views from search engines only, so a question asked of ChatGPT instead of Google does not appear in its denominator at all.
Management says the raise reflects faster delivery of capacity to meet demand already signed, and the backlog rose more than $50B in the same quarter (C09).
- 2026-10-29editedQ3 2026 call: the guide held, 2027 up significantly
Google says Argon ties OpenAI's GPT-6 Astra on a security benchmark and sets a high score on long engineering tasks (S12).
- 2026-10-28editedQ3 2026 release: the share program was used, $12.4B in one quarter
- 2026-10-29editedQ3 2026 10-Q wording of the programme's use
What we still do not know
- How much of the $513.9B of Cloud contracts depends on a few AI labs, which the filings do not name.
- How many questions people now ask chatbots instead of search engines; the firms that measure it sell the figure.
- What return Alphabet expects on its data-centre spending; it has given no target.
The evidence under this - 4 claims
Contracts of one year or less were added to the backlog definition from Q1 2026, so part of the growth is a change in what is counted.
- 2026-10-29editedQ3 2026 10-Q, filed after the first read: backlog $567.7B
The 10-Q does not split the backlog by customer, so how much rests on a few AI labs cannot be checked from the filing.
StatCounter counts page views from search engines only, so a question asked of ChatGPT instead of Google does not appear in its denominator at all.
Management says the raise reflects faster delivery of capacity to meet demand already signed, and the backlog rose more than $50B in the same quarter (C09).
- 2026-10-29editedQ3 2026 call: the guide held, 2027 up significantly
What would change our view
- Neutral→AttractiveThe price falls 20% below fair value
It needs:
- ○The share price falls to about $245, 20% below what we think the business is worth.about ≈$242.78 today, 13% below the last close
- Neutral→UnattractiveThe build keeps draining cash
It needs:
- ○Free cash flow over twelve months falls below zero, from $19.7B.checked at the next results, due by 15 Feb
- Neutral→UnattractiveCloud or search slows sharply
It needs any of:
- ○Google Cloud's sales growth falls below 40%, from 103%.checked at the next results, due by 15 Feb
- ○Search and other ad sales growth falls below 8%, from 12%.checked at the next results, due by 15 Feb
- Neutral→UnattractiveThe price runs 20% above fair value
It needs:
- ○The share price rises to about $368, 20% above what we think the business is worth.about ≈$374.46 today, 35% above the last close
We would also be wrong if
- Alphabet sells new shares beyond what it said the share programme is for, tax on staff stock pay.
- Free cash flow over twelve months is still below $20B at the end of 2027.
What is coming
- 9 November 2026 Q3 2026 Form 10-Q: The backlog, the purchase commitments and backstops, TPU agreements, and the stated use of the share program against the $12.4B sold in Q3.
Every dated event ahead
- 9 November 2026 Q3 2026 Form 10-Q: The backlog, the purchase commitments and backstops, TPU agreements, and the stated use of the share program against the $12.4B sold in Q3.
- December 2026 GFiber contribution to a new entity closes: Whether Alphabet keeps trimming Other Bets for cash while it funds the build; small in money, a signal on discipline.
- February 2027 Q4 2026 results, FY2026 10-K and the 2027 capital spending guide: How much 2027 spending rises, whether free cash flow recovers toward 2025's level, and whether buybacks restart or the share program is used.
- 15 May 2029 Mandatory convertible preferred converts into shares: Adds 43M to 54M shares; the count at conversion shows whether buybacks ever returned to absorb it.
Since we wrote this
- 9 NovThesis reviewNeutral: priced near a value of about $315Re-valued in three stages at an 8% return, the method now drawn on every Attractive thesis: the price sits at about its base-case value.
- 3 NovPrice moveThe price fell through our line; view moved to AttractiveAn 18% fall since October put Alphabet at 22.2 times operating profit, below the 22.5 times that route P1 set, so the view moved from Neutral to Attractive.
- 2 NovNewsAlphabet opened a euro bond sale of about 6 billion euros to fund spending. View heldAlphabet opens a 6 billion euro bond sale.
- 29 OctEarningsThe call kept spending rising; Neutral heldManagement left 2026 spending unchanged, said 2027 rises significantly and blamed slower search on currency and cheaper AI Mode queries. Cloud contracts reached $560.8B.
- 28 OctEarningsCloud doubled and met the bar; cash and shares worsenedCloud +103% at a 36.2% margin and search +12% passed our test, but free cash flow fell to negative $9.1B and $12.4B of shares were sold, so Neutral held.
- 27 OctNewsGoogle will offer its TPU chips inside Oracle's cloud from next year. View heldGoogle TPUs to be offered on Oracle Cloud from next year.
- 21 OctNewsOpenAI will sell its models through Google Cloud, a new route to enterprise customers. View heldOpenAI to sell its models on Google Cloud.
- 5 OctThesis reviewThe long case now sits under the verdictAfter the five-name pilot, the owner asked for the long case on every thesis.
- 4 OctThesis writtenNeutral, medium convictionWritten three weeks before Q3 2026 results, after the summer's capital spending raise, the Gemini delays and the June share sale, with the shares at $340.35.
- 4 OctThesis reviewOwner's checklist: revenue streams, cost of growth, management, catalystsThe owner's checklist found the method silent on who runs the company, what growth costs, where the money comes from and what is dated ahead; this pass fills those four blocks from the proxy and filings.
Also checked 7 times with nothing new to act on.
Sources 23
Sources - 23
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-02-05Item 1 (products, segments, competition by class, workforce), revenue by type, operating income, share counts at both year endsbacksC01C16C35C47C48C51
- S02primarycompany release2026-07-22revenue by line, segment operating income, other income gain, EPS, capital spending, free cash flow, cash, debt, headcountbacksC02C04C05C06C07C11C17C31C35C38C39C42
- S03primarycompany filing2026-07-23revenue backlog, TPU system agreements, SpaceX holdings, Wiz, the June share sale, convertible preferred and ATM, debt issues, commitments and backstops, repurchases, paid clicks, legal mattersbacksC02C03C04C06C08C10C13C14C15C16C17C24C25C26C36C43C48C49
- S04primarytranscript2026-07-22Q2 2026 earnings call transcriptdata/cache/transcripts/GOOG-2026Q2.jsonthe capital spending guide, backlog growth, customers exceeding commitments, supply constraints, third-party capacity, AI usage, the coding gap, funding plans; analysts from Morgan Stanley, JPMorgan, Goldman Sachs, Bernstein and others
We are updating our full year 2026 CapEx guidance range to $195 billion to $205 billion
backsC09C10C12C14C19C28C43 - S05primarycompany release2026-02-042025 capital spending, free cash flow and repurchases, the first 2026 capital spending guide, the Waymo roundbacksC05C11C12C16C31C42
- S06secondaryvendor data2026-10-04Research pack (FMP market tier and filed statements), frozen at content/GOOG/research-pack.jsoncontent/GOOG/research-pack.jsonprice, market value, enterprise value, returns and their decomposition, the filed cohort's multiples, valuation history, consensus, short interest, earnings history, filed quarterly operating income, headcount seriesbacksC05C07C20C29C30C31C32C33C35C38C39C51
- S07primaryindustry data2026-10-01Google and Bing shares of worldwide search engine use by month
Search Engine Market Share Worldwide - September 2026 Google 89.94 % bing 5.29 %
backsC18 - S08primaryindustry data2026-07-30cloud infrastructure market size, growth and shares by provider; generative AI services growth; the fastest-growing second tier
Q2’s worldwide market share was 28%, 20%, and 15%, respectively
backsC22 - S09primaryindustry data2026-04-16US digital advertising by format: total, search, social and video, with growth rates
Search revenues (including AI search) continue to hold the largest share of revenue dollars, reaching $114.2 billion in 2025.
backsC21 - S10secondarynews2025-05-07Eddy Cue's testimony on falling Safari searches, Apple's plan to add AI search providers, Google's reply
Cue also attributed a first-time drop in Safari searches to AI's growing influence as a search engine alternative.
backsC20C38 - S11secondarynews2026-09-02Business Insider: Google yet again avoided a breakup of its business, but it will have to play nicer US District Court for the Eastern District of Virginiathe ad tech remedy ruling, the April 2025 liability finding, the 2025 rejection of a Chrome sale
A federal judge ruled Wednesday that Google doesn't have to break up its adtech business
backsC24C25C39 - S12secondarynews2026-09-30the Gemini 4 Argon release, its limited availability, the abandoned Gemini 3.5 Pro, the benchmark tie with GPT-6 Astra
it has fallen behind the frontier to rival models from OpenAI and Anthropic in areas such as engineering and cybersecurity
backsC27 - S13secondarynews2026-07-21the Gemini 3.5 Pro delay and the cheaper Flash models released instead
Google said that Gemini 3.5 Pro , the new frontier model it originally promised for June, is still in testing.
backsC27 - S14secondaryother2026-08-27the investigation and the events it citesbacksC37
- S15primarycompany filing2026-10-04open-market sales and purchases, the Rule 10b5-1 flag on eachbacksC34
- S16primarycompany filing2026-01-29Meta's self-service advertising auction, the rival product for advertisers' budgetsbacksC23
- S17primarycompany filing2026-07-29Azure cloud services and Bing search, the rival products in both halves of AlphabetbacksC23
- S18primarycompany filing2026-02-06AWS on-demand computing and sponsored advertising for sellersbacksC23
- S19primaryregulator2026-09-15short interest in GOOG and its share of floatbacksC33
- S20primarycompany filing2026-04-24directors' and officers' biographies, beneficial ownership and voting power as of April 6, 2026, the CD&A (the end of the senior executives' bonus, the TSR PSUs, the CEO's 2026 award with Waymo and Wing units), the 200M-share stock plan increasebacksC44C45C46
- S21primarycompany release2026-10-28revenue and operating income by segment, the cash flow statement, free cash flow, the balance sheet, employees and the share programme in Q3 2026backsC02C04C05C11C13C14C16C17C29C30C33C35
- S22primarytranscript2026-10-28Q3 2026 earnings call transcript (FMP), held 2026-10-28data/cache/transcripts/GOOG-2026Q3.jsonthe capital spending guide, why search growth slowed, TPU system sales and the inventory build, third-party capacity, Cloud backlog; analysts from JPMorgan, Morgan Stanley, Goldman Sachs, Barclays, Bank of America, AllianceBernstein and MoffettNathanson
Our full year 2026 CapEx guidance range of $195 billion to $205 billion remain unchanged. We continue to expect our CapEx to increase significantly in 2027
backsC12C50 - S23primarycompany filing2026-10-29revenue backlog and the stated use of the at-the-market share programme
As of September 30, 2026, we had $ 567.7 billion of remaining performance obligations (“revenue backlog”), of which $ 560.8 billion related to Google Cloud.
backsC08C13C29C30C51