AI Selloff, Explained 5 Ways: Capex Beat Cash by $26.4B

Five explanations for this week’s fall in AI stocks run through the headlines and the most-viewed posts on X: spending without a cash payoff, AI eating search, a memory-cost squeeze, policy risk before the US midterms, and debt. We set each one against a single number from the companies’ filings or from our earlier reporting.

The most common explanation, AI spending without a cash payoff, has the clearest test. Alphabet, Amazon, Microsoft and Meta reported about $175.5 billion of operating cash flow for the quarter to 30 September and about $201.9 billion of capital expenditures, so capex exceeded operating cash by about $26.4 billion, by our arithmetic. Microsoft was the only one of the four whose operating cash covered its capex.

Short explainer · AI Selloff, Explained 5 Ways: Capex Beat Cash by $26.4B

A vertical news Short narrated from this piece’s headline and opening line.

Business Pill · Free Cash Flow

An evergreen Business Engineer explainer on free cash flow. It explains the general idea only and does not discuss this week’s results or the sell-off.

Reuters: Nasdaq 100 set for a $2 trillion October loss (30 Oct). Reuters reported on 30 October that more than $2 trillion in market value is set to be wiped from the Nasdaq 100 this month, as Big Tech earnings intensified scrutiny of whether AI spending can generate returns quickly enough. That is a one-month change for the whole Nasdaq 100; our figure in this piece measures how far the Magnificent 7 sit below their record highs, so the two are different measures and are not compared.

Checking the mixed signals (30 Oct). The market account @HyperTechInvest listed what it calls mixed signals and concluded that the market does not trust the sustainability of AI demand. Several items match figures our pieces took from company filings: Kioxia guided October-to-December revenue up 0.8% on the quarter, which the post rounds to 1% (Kioxia Hits SK Hynix’s 80% Margin but Guides Revenue Flat); Intel’s Data Center and AI unit grew 78% to $7.3 billion, per its release (Intel’s AI Unit Earns $3B); Alphabet’s release says Google Cloud revenues grew 103% (Google Cloud Added $15.6B in a Year); and across Alphabet, Amazon, Microsoft and Meta, third-quarter capex exceeded operating cash flow by about $26.4 billion, with only Microsoft’s cash covering its capex (Capex Outran Cash). The other items in the list are the account’s own reading and are not checked here.

How the Headlines Framed It

The Financial Times headlined “Big Tech sheds nearly $900bn in rout over soaring AI spending” at 10:00 GMT on 30 October. Bloomberg’s 10:30 GMT headline read “Tech Stocks Poised to Drop After Amazon Adds to AI Spending Fear”, and CNN’s at 12:08 GMT read “Wall Street sell-off continues as investors question returns on massive AI spending”.

A Reuters markets story syndicated overnight carried the headline “Stocks slump, bonds leap as AI giants’ spending spooks investors”. Euronews used “Tech firms take fresh hit as disappointing earnings fan AI fears”, and the Wall Street Journal “AI Selloff Deepens, Oil Jumps and Crypto Falls”.

Others framed it around Google. The Economist’s headline read “Google’s $500bn bloodbath reflects the danger it faces from AI”, and NBC News ran “Google stock’s historically bad day raises fears of AI bubble popping”. Separately, a Reuters Breakingviews commentary on 29 October was headlined “Artificial intelligence reaches its Napster moment”. We read these as headlines; several of the articles are paywalled.

What Was Posted on X

Posted by @exencial_RP: a BofA Bubble Risk Indicator chart. The post says the recent wobble in the AI trade and a historic rotation to defensive sectors have pushed the indicators for US Healthcare and Staples stocks towards bubble-risk levels.

@HyperTechInvest lists what it calls mixed signals in the market, including AWS sold out through 2027, Intel’s data center growth alongside weak guidance, Kioxia’s flat guide and harder neocloud financing, and concludes that the market doesn’t trust the sustainability of AI demand.

What X Amplified

The most-viewed post we found came from The Kobeissi Letter. It said “Four of the Magnificent 7 stocks are now in bear market territory” and that the seven had erased “a combined -$4 trillion in market cap since their highs”. It had about 1.23 million views when we read it at 13:40 GMT on 30 October.

Our count from 29 October closes found two of the seven, Alphabet and Meta, 20% or more below their highs, and the seven about $3.4 trillion below their highest closes in total, by our arithmetic.

A StockSavvyShay post calling it “an absolute bloodbath for stocks reporting earnings this afternoon”, led by Intel at -26%, had about 90,000 views at the same read. Our after-hours scoreboard has the timed figures.

1. Spending Without a Cash Payoff

This is the explanation in the FT, Bloomberg, CNN and Reuters headlines above. On X, Wall St Engine wrote that “These 9 companies alone are on track to spend around $1 trillion in capital expenditures this year”, in a post with about 24,000 views at our read.

The number that tests it is cash. Our compilation of the four companies’ earnings releases puts capex at about 1.8 times operating cash flow at Amazon, 121% at Alphabet and about 104% at Meta, against 68% at Microsoft. Microsoft’s operating cash exceeded its capex by about $18.6 billion, by our arithmetic.

A post by @pablomartimar gave the four companies’ capex as $203.0 billion, against our $201.9 billion; both put operating cash flow at $175.5 billion. Our capex figure is gross purchases of property and equipment and excludes finance leases.

2. AI Eating Search

The Economist’s headline points here. On X, The Kobeissi Letter quoted Reddit’s chief executive: “AI mode does not drive traffic. Not for us and not to anybody as far as I know.” That post had about 270,000 views at our read.

The test is where Google’s own advertising grew. Alphabet’s release shows Google Search & other revenue up $6,783 million year on year, while Google Network, the ads Google places on partner properties, rose $101 million. As we read those lines, the pressure in this quarter’s filings shows on the network side; Google Search revenue still grew.

3. A Memory-Cost Squeeze

The memory-cost explanation comes mainly from X. Wall St Engine posted that Bank of America “removes Apple $AAPL from its U.S. 1 List” and quoted the bank’s note: “We lower our F27 iPhone est. by 9mn units to 247mn (vs. St at 260mn).” By our arithmetic in our BofA piece, 247 million is 5.0% below the Street’s 260 million.

The chip headlines on 30 October framed the sell-off around Intel’s outlook and PC weakness rather than memory costs. A headline listed on TradingView at 11:23 GMT read “Intel warning sparks chip rout as PC weakness clouds turnaround”, and Reuters at 13:25 GMT ran “Intel plunges 25%, triggers global chip rout on bleak outlook”. Our piece on Asia’s chip stocks follows that sell-off across exchanges.

4. Policy Risk Before the Midterms

Bloomberg’s 11:00 GMT headline read “AI Policy Fears Grip Stock, Bond Traders With Midterms Days Away”. The opening of that article, as we reported, says Wall Street expects Democrats to win at least one house of Congress, “setting the stage for a potentially volatile ride in the stock and bond markets with artificial intelligence oversight, defense spending and crypto legislation on the table”.

No company filing tests a policy explanation. The nearest figure in the posts we read is positioning: The Kobeissi Letter wrote that “Long speculative positioning in Nasdaq futures surged to +$55 billion as of October 20th, its highest on record”, in a 29 October post citing positioning as of 20 October, before the results. We did not check that figure against exchange data.

5. Debt and Credit

The Reuters headline pairs the stock slump with “bonds leap”. On X, @pablomartimar wrote that “Meta is selling stock, Amazon’s long-term debt went from $66B to $173B”; we have not checked those figures against the filings.

Among our own filings-first pieces, the debt example is Microsoft’s: by our arithmetic, a sale of at least $25 billion of notes would equal about 70% of the $35.5 billion of total debt in its prospectus supplement.

We add Oracle as a further example from our coverage. It set a target of at least $30 billion of free cash flow in fiscal 2029, after negative $23.7 billion in fiscal 2026. Its president of Cloud Infrastructure, Mahesh Thiagarajan, said of its Project Jupiter build: “We’re actually not too worried about what we’re doing because I think we’re building global infrastructure”, according to the Times of India.

The Structural Read

As we read the week’s headlines, most of them explained the sell-off through AI spending, and that is also the explanation the four companies’ own cash-flow lines can test most directly.

As we read Alphabet’s release, the search explanation meets a mixed record: Google Search & other revenue grew $6,783 million while Google Network grew $101 million.

As we read the posts, the most-viewed claims ran ahead of our counts: The Kobeissi Letter put four of the Magnificent 7 in bear-market territory, against our two, and $4 trillion of losses, against our $3.4 trillion below highs.

Three Implications

FOR THE SPENDING EXPLANATION Our compilation puts Amazon’s quarterly capex at about 1.8 times its operating cash flow and Microsoft’s at 68%, so the four companies sit on different sides of the cash line.

FOR THE SEARCH EXPLANATION Alphabet’s release shows Search & other up $6,783 million year on year and Google Network up $101 million, with the 10-Q recording a decrease in AdSense revenues.

FOR THE DEBT EXPLANATION Our Microsoft piece puts a sale of at least $25 billion of notes at about 70% of the $35.5 billion of total debt in its prospectus supplement.

The Business Engineer Lens

This story maps onto the Business Engineer framework The AI Capex Map.

The framework puts it this way: “The $1T capex cycle is being underwritten by a $1T concentrated demand commitment from two counterparties whose combined free cash flow is deeply negative.”

As we read it, this week’s filings show the first half of that sentence in quarterly cash: three of the four companies in our compilation, Amazon, Alphabet and Meta, put more into capex than their operations brought in. The framework was published before this week’s results and does not discuss the sell-off or the posts quoted here.

What Is Not Established

Headlines and posts give explanations; they do not show why any stock moved. We read most of the outlets’ stories as headlines only, and the X posts are their authors’ views. The filing figures cover one quarter, and our capex totals exclude finance leases.

The Bank of America figures are as quoted by Wall St Engine; we have not read the note. The positioning figure is The Kobeissi Letter’s, and the explanations are not ranked by how much each moved prices. Our earlier piece on Google’s drop unpacks the one-day move.

The Bottom Line

Of the five explanations in the week’s headlines and posts, the spending one is tested here with a four-company filing total: Alphabet, Amazon, Microsoft and Meta put about $26.4 billion more into capex than their operations brought in during the quarter, by our arithmetic, and only Microsoft’s operating cash covered its own build. The search, memory and debt explanations each have a filing or analyst figure in our earlier pieces, and the policy explanation has none in company filings.