Market value after the earnings reports was the day’s biggest thread on 1 November 2026, by our grouping of the news. Meta, Alphabet, Microsoft and Amazon lost about $665 billion of market value in the first trading session after each released its third-quarter results, by our arithmetic from daily closing prices and the share counts on their 10-Q filings, and Alphabet accounted for $496.5 billion of it, about 75%.
Alphabet and Meta disclose up to $133.4 billion of contingent support for the companies that build and power their data centers, by our arithmetic from their 10-Qs.
The threads below are market value after earnings; financing off the balance sheet; what the filings say; power and compute; models, agents and leaks; money and politics; and deals and security. Each item links to the full piece, which names its sources.
Business Pill · FREE CASH FLOW
An evergreen Business Engineer explainer on free cash flow, the cash a business has left after its capital spending. It explains the general idea and does not discuss the companies in this roundup.
AI Daily Briefing · video
Today’s stories in about a minute (1 Nov 2026).
The key insight: As we read it, the day’s coverage set the market’s verdict on the four biggest spenders next to what their filings commit them to: by our arithmetic they lost about $665 billion of market value in their first sessions after reporting, while Alphabet and Meta disclose up to $133.4 billion of contingent support for their data center partners, also by our arithmetic.
Meta lost $124.4 billion, Amazon $37.2 billion and Microsoft $6.5 billion in their first sessions after reporting, by our arithmetic, so Alphabet’s $496.5 billion was more than the other three combined. Read: the post-earnings loss.
The world’s ten most valuable companies were worth $28.14 trillion at the close on Friday, 30 October, down $1.52 trillion from $29.66 trillion at the end of September, by our arithmetic; thirty large AI-exposed stocks lost a combined $3.17 trillion of market value in October, and only four of the 30 rose. Read: the top ten and the AI stock slide.
Nvidia was the more valuable of the two on 204 of the 209 trading days this year up to 30 October, by our arithmetic, until Apple passed it on Friday; at Friday’s close Apple was worth $5.03 trillion and Nvidia $4.93 trillion, AppleInsider reported. Read: Apple and Nvidia.
Google Search & other brought in $63,350 million of Alphabet’s $127,577 million in revenue in the quarter ended 30 September 2026, according to Alphabet’s third-quarter earnings release; by our arithmetic that is 49.7%, the first quarter below half in the 32 quarters of figures we read. Read: Search’s share.
The cost to protect against a debt default at Amazon, Alphabet, Microsoft, Meta and Oracle widened over the past month to the highest levels going back to 2021, The Wall Street Journal reported, citing the bond data provider Solve; a Financial Times chart built on Bloomberg consensus estimates shows the combined free cash flow of Alphabet, Meta, Microsoft and Amazon falling to about $60 billion in 2027 and then rising to about $640 billion by 2030, values we read approximately from the chart. Read: the debt insurance and the cash-flow consensus.
Gartner’s Hype Cycle for Artificial Intelligence, 2026 places generative AI in the Trough of Disillusionment; by our arithmetic on Federal Reserve data, the Nasdaq Composite fell 77.9% from its March 2000 closing high and did not close above that level again until April 2015. Read: Gartner’s trough.
Meta, Alphabet, Microsoft and Amazon lost about $665 billion of market value in the first trading session after each released its third-quarter results, by our arithmetic from daily closing prices and the share counts on their 10-Q filings; Alphabet accounted for $496.5 billion of it, and Meta, Amazon and Microsoft for $124.4 billion, $37.2 billion and $6.5 billion, which add up to $168.1 billion by our arithmetic.
Financing Off the Balance Sheet
Alphabet’s $91.9 billion of data center backstops and power-equipment guarantees and Meta’s residual value guarantees make up the $133.4 billion, by our arithmetic; the latest 10-Qs of Microsoft, Amazon and Oracle disclose no comparable arrangement in a full-text search. Read: the partner guarantees.
A chart credited to ZeroHedge puts eight companies’ disclosed off-balance-sheet commitments at $4.11 trillion against our own count of $3.94 trillion, and line by line the two counts agree within $1 billion for five of the eight companies, by our arithmetic. Read: the two off-book counts.
When Meta put its Hyperion campus into a joint venture in October 2025, it contributed $4.30 billion of held-for-sale assets, net of liabilities, and received a one-time distribution of $2.55 billion, according to its 2025 annual report, about 59% returned in cash by our arithmetic; the $27.3 billion bond that financed Hyperion carries a 6.581% coupon, IFR reported, above the highest stated rate on any of Meta’s own notes. Read: Hyperion’s cash and the Hyperion bond.
A Bloomberg Intelligence chart shared on X projects a gross AI funding gap of about $330 billion in fiscal 2027, more than double the roughly $150 billion it shows for fiscal 2026, and Amazon, Alphabet and Microsoft reported a combined $1,828.7 billion of signed revenue not yet recognised at 30 September 2026, by our arithmetic from their 10-Qs. Read: the funding gap and the cloud backlog.
Investor Chamath Palihapitiya said on the All-In podcast that the money AI firms need to raise through the end of 2027 is “30% of all of what the Treasury’s about to issue in that same period”. Read: the AI debt comparison.
What the Filings Say
Anthropic’s S-1 lists 88 risk factors in about 66,225 words of body text, by our count, and the longest single one, at 4,066 words or 6.1%, is about the inputs behind compute: chips, electrical power and data-centre capacity. Read: the S-1’s risk factors.
Alphabet’s quarterly report, filed with the SEC on 29 October 2026, says: “During model testing, we have sometimes observed unexpected behavior, including autonomous agents taking actions outside of their intended scope”; across nine filings split into 382 risk factors, and leaving aside a catch-all Other group, cyber, safety and outage risks get the most words in 4 of the 9, by our count. Read: Alphabet’s new sentence and the cross-company count.
Power and Compute
Chamath Palihapitiya wrote on X that “Power is THE binding constraint in AI/SI. Few really have it.” If Elon Musk’s “around 1.2M Blackwells at Colossus 2” all sat in GB300 NVL72 racks at HPE’s published spec, they would draw about 2.2 GW of rack power, by our arithmetic, before cooling. Read: Chamath on power and the Colossus 2 power math.
Data centers in space could cost up to 100 times as much as those on Earth, Google’s head of data center energy said this week, The Information reported; Epoch AI now estimates that OpenAI’s Stargate site in Abilene holds 1,018,696 H100-equivalents of compute, up from 764,022, a 33% increase by our arithmetic. Read: space data centers and Abilene.
Nvidia’s Vera Rubin NVL72 produced 9.11 times as many tokens per dollar as the GB300 NVL72 at 200 tokens per second per user, and 20.9 times as many at 325, on an official preview of SemiAnalysis’s InferenceX benchmark. Read: Rubin’s benchmark.
South Korea exported $56.44 billion of semiconductors in October 2026, up 259.4% from a year earlier, according to the Ministry of Trade, Industry and Energy, and ASMPT booked a record HK$8.85 billion of new orders in the third quarter, its results release says. Read: Korea’s chip exports and ASMPT.
Alphabet, Amazon, Meta and Microsoft have guided 2026 capital spending that adds up to about $730 billion at the midpoints, against about $615 billion of US energy investment on the IEA’s figures; Bloomberg reported that TSMC is spending $265 billion to build fabrication plants in Phoenix. Read: capex and energy and TSMC’s Phoenix campus.
Models, Agents and Leaks
A model labelled gpt-6.2-astra appears in Codex usage screenshots posted on X, and OpenAI has not announced a GPT-6.2 model; a screenshot shows “Claude Mythos 5.5” in the code of Anthropic’s Mythos page, but when we fetched the page both places read “Claude Mythos 5.1”. Read: GPT-6.2 Astra and Mythos 5.5.
A Hugging Face upload presented as a leaked Anthropic checkpoint contains 43 tensors that can be rebuilt exactly from their own names with a seeded random number generator, according to an audit on the repository’s community page; Anthropic will start applying its text watermark to four older Claude models on 3 November 2026, according to a customer email shared on X. Read: the Hugging Face upload and the watermark.
Elon Musk said on X that “The cost of serving Grok Bot was unsustainably high relative to revenue”, and a Claude user says Anthropic flagged the account for “training a competitor”. Read: Grok Bot’s cost and Anthropic’s terms.
Revolut moved 2,000 engineers off metered AI APIs onto self-hosted open models running on 56 Nvidia B200 GPUs and says that cut its AI coding costs 90%, StartupHub.ai reported, and a FAIR at Meta paper puts a shared memory for Claude Opus 5 at $914.36 for a 398-task run against $161.77 with no memory. Read: Revolut and the memory cost.
Google’s FlowAgent took a median of 9.85 minutes from a failing-test notification to a successful fix, according to a Google paper on arXiv, and Meta runs Nvidia’s B200 GPUs at a 750-watt power limit so that it can cool them with air, according to a paper by Meta and AMD engineers. Read: FlowAgent and Meta’s B200 limit.
DeepSeek published dsh-superagent on GitHub, and Gemini stopped adding the tracking tag utm_source=gemini to the links it cites, according to server logs published by search consultant Natzir Turrado; Andrej Karpathy says Opus 5.5 built HackerTube, a site of 6,454 videos Hacker News users recommended, in about two hours. Read: DeepSeek’s agent, Gemini’s links and HackerTube.
Money and Politics
OpenAI plans to argue that the infrastructure it builds should be called “AI compute campuses” rather than data centers, Axios reported; if every candidate pledging a data center pause won on 3 November, the states where they are running would hold roughly 30 GW of capacity planned for 2028, according to an analysis by SemiAnalysis. Read: the rebrand and the pause pledges.
A New York Times list of the 20 biggest midterm donors puts Elon Musk first at $445.6 million, and donors from tech, AI and crypto account for $835.4 million, or 43.0%, of the list’s total, by our arithmetic; the Washington Post found that chatbots failed to say which congressional district an address was in 48 percent of the time. Read: the donor list and the voter guides.
Aramco has taken a 49% stake in HUMAIN by transferring AI assets and contributing about $2.6 billion in cash, the Public Investment Fund said; the Quad’s foreign ministers said they intend to provide “over $85 million of technical and financial assistance to support access to trusted SI across the Indo-Pacific”; and the UK’s Sovereign AI Fund is among the new investors in Fractile, The Telegraph reported. Read: Aramco and HUMAIN, the Quad and Fractile.
Elon Musk and Sergey Brin both said OpenAI president Greg Brockman had asked them for compute, according to a Colossus feature. Read: Brockman’s ask.
Deals and Security
IREN completed its acquisition of Mirantis on 4 August, according to its filing, and ServiceNow chief executive Bill McDermott said its Sweep acquisition moves customers between CRM systems in “days, 30 days, 60 days for a serious company of serious scale”, CTech reported. Read: the neoclouds’ software moves and Sweep.
Deepgram said it is acquiring Daily, the maker of Pipecat; Lumaril raised a $66 million Series A at a $400 million valuation, The Wall Street Journal reported; and KB Kookmin Bank said it will lead the financing for a Brookfield AI computing project worth about 12.6 trillion won, which it calls Korea’s first GPU financing. Read: Deepgram, Lumaril and the GPU loan.
A crypto perpetual contract on Anthropic traded at $111.10, about 1.88 times the Series H share price implied by PitchBook’s figures, by our arithmetic; OpenAI is seeking money at a $1.25 trillion valuation about 11 years after it was founded, according to the Financial Times and The Wall Street Journal. Read: the Anthropic perp and OpenAI’s valuation.
HD Hyundai Robotics said AI agents built with Palantir cut the average time to resolve robot-related problems from 21.5 days to 7.9 days, and EPB Research said three of the industries that use AI most lifted output per payroll job by 5.0% a year since the end of 2022. Read: HD Hyundai and the productivity data.
The FBI said North Korea was responsible for the theft of “more than $6 billion USD in virtual assets” from Binance, and Double Counter says data tied to about 28M Discord accounts is treated as exposed, according to an incident report published by Tellter. Read: the Binance alert and Double Counter.
Approximate combined free cash flow of Alphabet, Meta, Microsoft and Amazon, as our piece reads it from a Financial Times chart built on S&P Capital IQ figures and Bloomberg consensus estimates: about $230 billion in 2024, about $60 billion in 2027 and about $640 billion in 2030.
The Structural Read
By our arithmetic, the ten most valuable companies were worth $28.14 trillion at the close on 30 October, down $1.52 trillion from the end of September, and Alphabet lost $780.2 billion, about 51% of the ten’s combined fall.
A chart credited to ZeroHedge puts eight companies’ disclosed off-balance-sheet commitments at $4.11 trillion, against our own count of $3.94 trillion, and the bond that financed Meta’s Hyperion campus carries a 6.581% coupon, IFR reported.
As we read it, power ran through the compute coverage: Anthropic’s S-1 gives its longest risk factor, 4,066 words, to chips, electrical power and data-centre capacity, by our count, and Musk’s 1.2M Blackwells would draw about 2.2 GW of rack power before cooling, by our arithmetic.
Chamath Palihapitiya, on X, 1 November 2026, as quoted in our piece
“Power is THE binding constraint in AI/SI. Few really have it.”
Three Implications
THE FOUR BIGGEST SPENDERS By our arithmetic, Meta, Alphabet, Microsoft and Amazon lost about $665 billion of market value in the first session after their third-quarter reports, and a Financial Times chart built on Bloomberg consensus estimates shows their combined free cash flow at about $60 billion in 2027, values we read approximately.
LENDERS AND PARTNERS The cost to protect against a debt default at Amazon, Alphabet, Microsoft, Meta and Oracle widened to the highest levels going back to 2021, The Wall Street Journal reported, citing Solve, and Alphabet and Meta disclose up to $133.4 billion of contingent support for their data center partners, by our arithmetic.
LABS AND MODEL BUYERS Elon Musk said on X that the cost of serving Grok Bot was unsustainably high relative to revenue, and Revolut says self-hosted open models cut its AI coding costs 90%, StartupHub.ai reported.
The analysis says: “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, the day’s coverage shows the market pricing that cycle, in the four companies’ post-earnings losses, the debt insurance and the commitments kept off balance sheets. We do not say one caused the other. The analysis was published in May 2026 and does not discuss these events.
What Is Not Established
This roundup adds no new reporting and selects from the day’s coverage. Every figure above is taken from the linked FourWeekMBA piece, each built from its own primary source. The seven threads are our grouping, not a finding.
We do not claim that these stories are connected or that one caused another. The market-value losses, the off-book count and the backlog are our arithmetic from closing prices and company filings; the debt-insurance levels rest on The Wall Street Journal citing Solve, the cash-flow path on our approximate reading of an FT chart of Bloomberg consensus, the funding gap on a Bloomberg Intelligence chart shared on X, and the GPT-6.2 and Mythos 5.5 traces on screenshots posted on X.
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 daily closing prices and 10-Q share counts, Meta, Alphabet, Microsoft and Amazon lost about $665 billion of market value in the first session after their third-quarter reports, and Alphabet’s $496.5 billion was about 75% of it. Their 10-Q filings for the quarter to 30 September show Alphabet and Meta disclosing up to $133.4 billion of contingent support for their data center partners, by our arithmetic.
This week’s AI Supercycle issue on The Business Engineer reads the build-out as a credit story: read the issue.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. This roundup draws on FourWeekMBA’s news coverage of 31 October and 1 November 2026, linked above; every story was built from its own primary source, which it cites and links. We re-read each piece on 1 November 2026 and took every figure here from it. The seven threads and the figures in our charts are our own grouping and selection. Nothing here is a forecast, and nothing here is financial or investment advice.
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.
