The five biggest AI builders, Amazon, Alphabet, Microsoft, Meta and Oracle, spent about $31.8 billion more on property and equipment in their latest quarter than their operations brought in, by our arithmetic from their 10-Q filings. A year earlier the same five had about $61.4 billion left over, so the swing is about $93.2 billion.
A Goldman Sachs chart shared on X by Mike Zaccardi puts the five at minus $31 billion for the third quarter and shows every quarter since the start of 2024 above zero until this one.
Short explainer · Big Tech free cash flow hit -$31.8B
A vertical news Short summarising this story. Figures come from the five companies’ 10-Qs and Goldman Sachs’s exhibits as posted on X; our arithmetic is labelled.
Business Pill · Free Cash Flow
An evergreen Business Engineer explainer on free cash flow. It teaches the general idea only and does not discuss any company in this piece.
The key insight: As we read the filings, the five biggest AI builders added about $124.6 billion of quarterly capital spending in a year while their operating cash rose about $31.5 billion, which is how a $61.4 billion surplus became a $31.8 billion gap, by our arithmetic.
The exhibit, posted by Zaccardi on X, is titled “Hyperscaler aggregate free cash flow turned negative in Q3” and lists AMZN, GOOGL, META, MSFT and ORCL. Its source line reads “FactSet, Goldman Sachs Global Investment Research”.
The only value labelled on the chart is minus $31 billion for the third quarter of 2026. The bars for the first and second quarters of 2026 are small and positive, at roughly $9 billion and $6 billion as read from the chart, against roughly $62 billion in the third quarter of 2025. We did not read Goldman’s note.
ZeroHedge headlined its write-up “Hyperscalers Burn $31 Billion In Q3 As Goldman Warns The AI Earnings Boom Has Stopped Accelerating”. Its article is behind a paywall and we did not read it.
Operating cash flow minus purchases of property and equipment in the third quarter of 2025 and 2026 for Amazon, Alphabet, Microsoft, Meta and Oracle (Oracle: fiscal first quarter, June to August), from their 10-Qs. Totals are FourWeekMBA arithmetic.
What Was Posted on X
Animated chart: 10-Q arithmetic for Q3 2025 and Q3 2026; other quarters read approximately from Goldman Sachs Exhibit 16 (FactSet).
— Mike Zaccardi, CFA, CMT ๐ (@mikezaccardi) Oct 31 14:16
Mike Zaccardi posted Goldman Sachs’s Exhibit 16, titled “Hyperscaler aggregate free cash flow turned negative in Q3”, covering Amazon, Alphabet, Meta, Microsoft and Oracle, with FactSet as the data source. Its only labelled value, minus $31 billion for the third quarter, is the figure this piece checks against the companies’ 10-Qs.
GS: Consensus estimates show hyperscaler free cash flow declining further in 2027 but rebounding in 2028
— Mike Zaccardi, CFA, CMT ๐ (@mikezaccardi) Oct 31 15:22
Zaccardi also posted Goldman’s Exhibit 17, titled “Consensus estimates show hyperscaler free cash flow declining further in 2027 but rebounding in 2028”, which charts realized and consensus free cash flow for the same five companies from 2022 to 2028. The piece reads its annual values approximately from the chart.
What the Filings Show
We rebuilt the number from each company’s 10-Q, taking operating cash flow minus purchases of property and equipment for the quarter to 30 September. Oracle’s matching quarter is its fiscal first quarter, June to August.
Amazon was minus $34.8 billion, with $77.3 billion of capital spending against $42.4 billion of operating cash. Alphabet was minus $9.1 billion, Oracle minus $5.4 billion and Meta minus $1.2 billion. Microsoft was the only one in surplus, at plus $18.6 billion.
Together that is minus $31.8 billion, close to the minus $31 billion on Goldman’s chart, by our arithmetic. In the same quarter of 2025 the five were at plus $61.4 billion, with only Oracle slightly negative.
Spending Rose Four Times as Many Dollars as Cash
The five companies’ operating cash flow rose from $167.1 billion to $198.6 billion, up about $31.5 billion or 19%. Their capital spending rose from $105.8 billion to $230.4 billion, up about $124.6 billion or 118%, by our arithmetic.
In dollars, spending grew about four times as much as cash. Capital spending equalled about 116% of operating cash flow in the quarter, against about 63% a year earlier.
What Consensus Estimates Show for 2027
A second Goldman exhibit, also posted by Zaccardi, is titled “Consensus estimates show hyperscaler free cash flow declining further in 2027 but rebounding in 2028” and uses the same five companies and the same FactSet source.
As read from that chart, the five companies’ annual free cash flow was roughly $240 billion in 2024 and roughly $190 billion in 2025. The consensus estimates it plots are roughly minus $45 billion for 2026, roughly minus $155 billion for 2027 and roughly plus $60 billion for 2028. These are analysts’ estimates compiled by FactSet, not company guidance, and the values are approximate readings of the chart.
The Business Lever: Who Funds the Gap
When capital spending exceeds operating cash, the difference has to come from cash on hand, borrowing or other financing. This week’s coverage of the same companies includes $25 billion of Microsoft bonds, off-balance-sheet commitments of about $3.94 trillion and a 15-year low in cash returned to shareholders.
The Structural Read
As we read the 10-Qs, Amazon accounts for about $34.8 billion of the gap on its own, more than the five-company total, and Microsoft’s $18.6 billion surplus offsets part of the rest.
As we read Goldman’s first exhibit, every quarter from the start of 2024 to the second quarter of 2026 was positive, and the first two quarters of 2026 were already small.
As we read Goldman’s second exhibit, the consensus estimates it plots put 2027 lower than 2026 and 2028 back above zero; these are analysts’ estimates compiled by FactSet, read approximately from the chart.
Goldman Sachs Global Investment Research, Exhibit 17, as posted by Mike Zaccardi
“Consensus estimates show hyperscaler free cash flow declining further in 2027 but rebounding in 2028”
Three Implications
FOR THE COMPANIES The gap between capital spending and operating cash has to be met from cash on hand, borrowing or other financing.
FOR LENDERS AND INVESTORS This week’s coverage includes Microsoft’s $25 billion bond sale and about $3.94 trillion of off-balance-sheet commitments across Big Tech.
FOR THE MEASURE Our figures use operating cash flow minus purchases of property and equipment from the 10-Qs; Goldman’s FactSet measure may differ, and its chart reads minus $31 billion.
The framework sets out to map the financing of AI: “How the money flows, where it cracks, what it buys, who can absorb it”.
As we read it, the third quarter is the first in Goldman’s series where the five builders’ spending ran ahead of the cash their operations produced. The framework was published in May 2026 and does not discuss this quarter.
Our measure excludes finance-lease principal payments, so it differs from some companies’ own free-cash-flow figures and may differ from Goldman’s FactSet measure. The quarterly bar heights before the third quarter of 2026, and the annual figures in Goldman’s second chart, are read from the charts and are approximate; the 2026 to 2028 figures are consensus estimates, not results. Goldman’s series in the chart starts in 2024, so it does not show whether the five were negative in earlier years.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
By our arithmetic from their 10-Qs, the five biggest AI builders spent about $31.8 billion more than their operations generated in the third quarter, a swing of about $93.2 billion from a year earlier, and Goldman’s chart puts the figure at minus $31 billion.
A note on sourcing. Company figures come from the five companies’ 10-Q filings via SEC XBRL; Goldman Sachs’s two exhibits were read as posted on X by Mike Zaccardi on 31 October 2026, and Goldman’s note was not read. Totals and ratios are our arithmetic. We did not contact the companies.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
