For twenty years, the single most reliable fact about Alphabet was that it printed cash and handed a large slice of it back. In 2025 it repurchased $45.7 billion of its own stock. In the first half of 2026 it repurchased zero — and instead sold $30.5 billion of common stock, $19.1 billion of mandatory convertible preferred, and $24.8 billion of debt in a single quarter.
That is the fact that reframes every GOOG price target on the Street. Alphabet has stopped being a company that returns capital and become one that raises it. The buyback did not shrink; it stopped. And the timing produces an irony almost nobody has stated out loud: in the very quarter Berkshire Hathaway was buying roughly $17 billion of Alphabet shares, Alphabet itself was issuing $49.6 billion of new ones. The most famous value investor on earth was accumulating the stock while the company was manufacturing it.
With GOOG at $341.45 (17 August 2026 close), this piece lays out a $460 bull case and a $235 bear case, both built from Alphabet’s own cash-flow statement rather than from a headline EPS number that is, as we will show, 69% paper gains.
Quick Take: the numbers that matter
- Spot: $341.45 (17 Aug 2026 close), 14.4% below the 13 May closing high of $399.04 and 70.6% above the August 2025 low — StockAnalysis
- Q2 2026 revenue: $119.8bn, +24%, a 12th consecutive quarter of double-digit growth — Alphabet Q2 2026 release, 22 July 2026
- Google Cloud: revenue $24.8bn, +82%; operating income $8.8bn vs $2.8bn, margin 20.7% → 35.6%
- Reported EPS $9.11 — of which $6.26 came from unrealised gains on equity securities, per Alphabet’s own footnote. Operating EPS was $2.85
- Free cash flow by quarter: +$24.5bn → +$24.6bn → +$10.1bn → −$5.9bn (Q2 2026)
- Capex: $44.9bn in Q2 alone; 2026 guidance $195–205bn, 2027 consensus about $257bn per FactSet
- Buybacks: $28.3bn in H1 2025 → $0 in H1 2026
- Street: “Strong Buy”, average target $428.04 across 64 analysts; the single most bearish target is $340 — S&P Global via StockAnalysis
The capital reversal nobody is pricing
Go to page 7 of Alphabet’s Q2 release, the consolidated statement of cash flows, and read the financing section. It is the most changed page in the document.
| Financing line (Q2) | 2025 | 2026 |
|---|---|---|
| Repurchases of stock | −$13,238m | $0 |
| Proceeds from issuance of common stock | $0 | +$30,499m |
| Proceeds from mandatory convertible preferred | $0 | +$19,063m |
| Proceeds from issuance of debt, net | +$26,846m | +$24,847m |
| Net cash from financing | −$5,832m | +$61,243m |
Alphabet described the equity raise in its own words: it issued Class A stock, Class C stock and mandatory convertible preferred “for aggregate net proceeds of $49.6 billion, to be used for general corporate purposes, including capital expenditures to scale AI infrastructure and global compute.” It also put in place an at-the-market programme to sell up to a further $40.0 billion of stock, none of which had been drawn at 30 June.
A company swinging from returning $5.8 billion to absorbing $61.2 billion in a single quarter is not making a financing tweak. It is telling you the AI buildout has outgrown the cash the business produces. Alphabet’s long-term debt went from $46.5 billion at the end of December to $98.2 billion at the end of June — it more than doubled in six months.
Having watched a decade of megacap capital allocation, I cannot think of a comparable reversal at a company of this quality that the market has treated this calmly. GOOG is down only 14% from its high. Berkshire, meanwhile, added around 48.1 million shares in the quarter, making Alphabet one of its three largest equity positions. Both things are true at once, and reconciling them is the whole investment question.
The EPS illusion: $9.11 reported, $2.85 earned
Alphabet reported net income of $112.1 billion in Q2 and EPS of $9.11, up 294%. Those numbers are real in the GAAP sense and close to meaningless in the valuation sense.
Other income was +$97,983 million against $2,662 million a year earlier. Of that, $99,031 million was a gain on equity securities — Alphabet’s stakes in other companies being marked up. The footnote is unusually explicit: the $99.0 billion gain “increased the provision for income tax, net income, and diluted net income per common share by $21.9 billion, $77.1 billion, and $6.26, respectively.”
Strip it out and the picture is sober but healthy:
| Metric (Q2) | 2025 | 2026 | Change |
|---|---|---|---|
| Reported diluted EPS | $2.31 | $9.11 | +294% |
| Contribution from equity-securities gain | negligible | $6.26 | — |
| Operating EPS ex-gain | $2.31 | $2.85 | +23.4% |
| Operating income | $31,271m | $40,770m | +30% |
So the business grew earnings roughly 23–30%, not 294%. Annualise the clean $2.85 and Alphabet is on about $11.40 of operating EPS, which puts GOOG at $341.45 on roughly 30 times earnings — not the single-digit multiple the headline number implies, and not the 60-plus multiple a naive AI-basket comparison would suggest.
The trap runs the other way too, and this is the part worth writing down. Those equity stakes are marked both directions. Alphabet’s own footnote warns that “fluctuations in the value of our investments may be affected by market dynamics and other factors and could significantly contribute to the volatility of OI&E in future periods.” If the private-company valuations behind this markup compress, Alphabet will report a spectacular “loss” in a quarter where the search and cloud businesses did nothing wrong. Do not sell it, and do not let anyone tell you the earnings collapsed.
The free cash flow cliff, in Alphabet’s own table
Alphabet publishes its own free cash flow reconciliation, and the four-quarter sequence is the clearest thing in the release:
| Q3 2025 | Q4 2025 | Q1 2026 | Q2 2026 | |
|---|---|---|---|---|
| Operating cash flow | $48,414m | $52,402m | $45,790m | $39,069m |
| Capital expenditure | −$23,953m | −$27,851m | −$35,674m | −$44,924m |
| Free cash flow | +$24,461m | +$24,551m | +$10,116m | −$5,855m |
We flagged the turn when it happened, in our coverage of Alphabet’s first negative free cash flow quarter, and the market’s reaction was immediate: GOOG fell 6.89% on 23 July, part of the broad repricing of AI capital spending that also hit Tesla.
Here is the synthesis the individual numbers do not give you. Trailing-twelve-month operating cash flow is $185.7 billion. Management guides to $195–205 billion of capex in 2026, and consensus has 2027 at about $257 billion. Hold operating cash flow flat and 2027 capex alone would exceed everything the business generates by roughly $70 billion. Even growing operating cash flow 20% a year, Alphabet does not cover its own 2027 capital plan from operations.
That is not a cyclical dip in free cash flow. It is a structural decision to fund a decade of infrastructure from the capital markets — and the $40 billion at-the-market programme, still untouched, is the pre-positioned instrument for doing it. Note also that H1 capex was $80.6 billion against full-year guidance of $195–205 billion, which means second-half spending is set to run roughly 50% above the first half. The worst free cash flow quarters are ahead, not behind.
What is actually working, and why the bears keep losing
The consensus bear thesis for two years has been that generative AI erodes search. The Q2 numbers say the opposite. Google Search & other revenue grew 17% to $63.3 billion — an acceleration, not a decay. YouTube ads grew 13% to $11.1 billion. Google Services in aggregate grew 15%.
The bigger story is Cloud. Revenue rose 82% to $24.8 billion, and operating income more than tripled from $2.8 billion to $8.8 billion. That is a margin expansion from 20.7% to 35.6% while the segment is growing at 82% — the combination that almost never happens, because scaling businesses normally trade margin for growth. Sundar Pichai, CEO of Google and Alphabet, put the demand signal plainly: “Gemini models now process 22 billion API tokens per minute and the Gemini App has 950 million monthly active users.”
Two structural offsets matter for the capex bear case. First, Alphabet designs its own silicon — Cloud now books product revenue “primarily from the sale of TPU systems,” meaning some of that enormous capex converts into revenue from other people’s AI workloads rather than pure cost. Our reporting on Google’s Frozen v2 efficiency chip covers the economics of that vertical integration. Second, Alphabet is the only hyperscaler whose AI spending is partly self-supplied, which is why its capex buys more compute per dollar than a competitor buying merchant GPUs.
And Alphabet’s power over the open web is undiminished — arguably the opposite. Our analysis of how Reddit grew revenue 61% and still fell 9% because of Google is a useful case study in who actually holds the pricing power in the AI-search transition.
The pressures: talent, courts, and a widening cost centre
Three risks sit outside the financial statements.
The first is people. On 5 August, chief scientist Jeff Dean left after 27 years to co-found Discovery Loop, taking senior fellow Sanjay Ghemawat, DeepMind vice president Oriol Vinyals and Google Brain co-founder Quoc Le with him. On the same day, Demis Hassabis stepped down as CEO of Google DeepMind, moving to chairman of the unit and adding the role of chief scientist at Alphabet. The shares fell about 5% within minutes. Hassabis has not left the company, which several write-ups blurred — but losing four of the most cited researchers in the field in one announcement, at the precise moment you are committing $200 billion a year to infrastructure they helped justify, is a genuine risk to the return on that spend.
The second is the courts. Alphabet was ordered to pay Klarna’s PriceRunner $1.9 billion in damages — a private antitrust action, not a regulator’s fine, which is the more expensive category because it invites copycats with contingency-fee lawyers behind them.
The third is buried in the segment table. “Alphabet-level activities,” which the company says “primarily reflects expenses related to our shared AI research and development,” widened from a −$3,372 million quarterly loss to −$5,789 million. That unallocated AI cost centre is growing 72% year on year and sits outside both Services and Cloud, so it flatters neither segment’s reported margin while consuming real money. Other Bets lost a further $1,799 million on $382 million of revenue.
The $460 bull case
Target: $460 — a 35% gain from $341.45.
The bull case does not require the capex to stop. It requires the capex to be right, and it requires 2026 to be the peak year of the funding gap.
At $460, Alphabet is worth roughly $5.7 trillion on about 12.3 billion diluted shares. Against operating EPS of roughly $13.50 in 2027 — Q2’s clean $2.85 annualised and grown at the 23% rate the business is currently compounding at — that is about 34 times earnings. Rich, but not absurd for a company whose second segment is growing 82% with expanding margins, and comfortably beneath the multiples the market is paying for AI infrastructure names with a fraction of the cash generation.
Three things have to happen. Cloud growth must stay above 50% so the capex has visible revenue attached to it. Free cash flow has to turn positive again during 2027 as the 2026 build depreciates into a revenue base rather than a hole. And the at-the-market programme has to stay largely undrawn — if Alphabet funds the rest of the build from operating cash and debt rather than equity, the dilution overhang lifts and the stock re-rates on the same earnings. The Street’s $428.04 average target sits just below this scenario, which tells you the sell-side already assumes most of it.
The $235 bear case
Target: $235 — a 31% decline from $341.45.
The bear case is not that Google loses search. It is that the market decides to value Alphabet the way it values every other business that spends more than it earns on physical infrastructure.
At $235 the market capitalisation is about $2.9 trillion and the multiple on clean earnings is roughly 21 times — still a premium to the market, and well above the August 2025 low of $200.19 that this stock printed only twelve months ago. A 31% drawdown sounds dramatic; GOOG has moved more than that in the past year in the other direction, and realised volatility over the last 90 sessions is 37.7%.
The trigger is arithmetic. If 2027 capex lands near the $257 billion consensus and operating cash flow grows at anything less than a heroic rate, Alphabet issues equity again — this time into a market that has already been told the buyback is gone. Capital-intensive businesses get capital-intensive multiples, and the re-rating from 30 times to 21 times is exactly what that transition looks like. Add a mark-down cycle in those equity stakes, which would produce an ugly GAAP loss at the worst possible moment for sentiment, and the sequence is easy to imagine.
One detail should give even committed bulls pause: of 64 analysts covering the stock, the most bearish price target is $340 — essentially spot. Not one sell-side model on the Street contemplates GOOG falling. That is not a sign of safety. It is a sign that the downside case is unowned, which is precisely the condition under which downside moves are fast. Compare the sell-side’s behaviour with the more balanced positioning we found in our IonQ bull and bear analysis, where an equal-weight target sat below the average and the distribution had two sides.
Scenario table
| Scenario | Price | vs spot $341.45 | Market cap | Multiple on clean EPS | Requires |
|---|---|---|---|---|---|
| Bull | $460 | +35% | ~$5.7tn | ~34× 2027E | Cloud growth >50%; FCF positive in 2027; ATM stays undrawn |
| Street | $428.04 | +25% | ~$5.3tn | ~32× | Average of 64 analysts, “Strong Buy”; lowest target $340 |
| Bear | $235 | −31% | ~$2.9tn | ~21× | 2027 capex near $257bn; further equity issuance; multiple compresses to capital-intensive |
What happens next
First, watch the at-the-market programme, not the earnings. Alphabet has authorised $40 billion of stock sales and drawn none of it. The 10-Q that first shows material ATM usage is the single most important document GOOG holders will read this year — it converts the dilution risk from theoretical to actual and would confirm that operating cash flow plus debt is not covering the build.
Second, expect the buyback question to dominate Q3. Alphabet still pays a $0.22 quarterly common dividend and has now added a $12.15 quarterly preferred dividend on the Series A and Series B mandatory convertibles. A company paying preferred dividends while running negative free cash flow and not repurchasing shares has visibly reordered its capital priorities. Management will be asked when repurchases resume; the honest answer, on the current capex path, is not before the buildout peaks.
Third, treat the next reported EPS number as noise until you have read the footnote. Whatever Q3 prints, find the “gain (loss) on equity securities, net” line first and subtract its per-share effect. On this quarter that single adjustment moved EPS from $9.11 to $2.85. It will move it again, possibly in the other direction.
On balance the base case sits nearer the Street’s $428 than the bear case, because the operating business is performing better than the bears claimed and the balance sheet can absorb a great deal. But the honest summary is that Alphabet has swapped a known, boring, cash-returning profile for a leveraged bet on AI infrastructure demand, and it has done so while its most senior researchers walked out the door. The stock is priced as though only the first half of that sentence is true.
Frequently asked questions
What is the GOOG stock prediction for 2026?
Our scenarios put GOOG at $460 in the bull case and $235 in the bear case from a spot price of $341.45 on 17 August 2026. The Street’s average target is $428.04 across 64 analysts, with a “Strong Buy” consensus and a lowest target of $340. The spread reflects uncertainty about Alphabet’s capital spending, not about its revenue growth.
Why did Alphabet stop buying back its stock?
Alphabet repurchased $28.3 billion of stock in the first half of 2025 and nothing at all in the first half of 2026. The cash is going into AI infrastructure instead: capital expenditure guidance for 2026 is $195–205 billion. In Q2 2026 the company raised $49.6 billion in equity and $20.3 billion in senior unsecured notes rather than returning capital.
Is Alphabet’s free cash flow really negative?
Yes, for one quarter so far. Alphabet’s own reconciliation shows free cash flow of −$5,855 million in Q2 2026, after +$10,116 million in Q1 and +$24,551 million in Q4 2025. Trailing-twelve-month free cash flow is still positive at $53.3 billion, but with second-half capex guided roughly 50% above the first half, more negative quarters are likely.
Why was Alphabet’s Q2 EPS $9.11 if the business only grew 24%?
Because $6.26 of that $9.11 came from unrealised gains on Alphabet’s equity holdings, which the company disclosed in a footnote. Operating EPS excluding that gain was $2.85, up 23.4% from $2.31 a year earlier. Any valuation built on the reported $9.11 overstates Alphabet’s earnings power by roughly three times.
Is AI destroying Google Search revenue?
Not on the current numbers. Google Search & other revenue grew 17% year over year to $63.3 billion in Q2 2026, an acceleration rather than a decline. Google Cloud grew 82% to $24.8 billion with operating margin expanding from 20.7% to 35.6%. The pressure on Alphabet is showing up in cash flow, not in demand.
What is the biggest risk to Alphabet stock right now?
Dilution. Alphabet has a $40 billion at-the-market equity programme authorised and undrawn. If 2027 capital expenditure approaches the $257 billion consensus, the company will likely have to use it, issuing shares into a market that has already lost the buyback. That combination — more shares, no repurchases, negative free cash flow — is what would compress the multiple toward the bear case.
This article is for information purposes and does not constitute investment advice. Price data as of the 17 August 2026 close. Financial figures are taken from Alphabet’s Q2 2026 earnings release and SEC filings.