AI Daily: Nvidia at 13.3x Earnings, Below the S&P’s 18.3x

A day before its 17 November results, Nvidia traded at 13.3 times the analyst consensus for its next four quarters of earnings in morning trading on Monday, 16 November 2026, by our arithmetic, below the 18.3 times FactSet gives the S&P 500. CNBC reported that the board is not considering bringing back Jensen Huang as chief executive, and the 10-Q puts Nvidia’s guarantee exposure at $108.5 billion, against a Bank of America estimate of $230 billion.

Later on 16 November: Nvidia closed at $156.14, down 4.12% on the day, according to Nasdaq’s quote data, leaving it $423.9 billion of market value below its 11 November close, the last before the 8-K, by our arithmetic. At the close it traded at 13.0 times the $12.02 consensus for its next four quarters, by our arithmetic. Our piece on the close has the full numbers.

As we read it, the day’s coverage kept putting a price on AI’s biggest names before their big weeks: Nvidia below the index on expected earnings, Anthropic at 14.6 times run-rate revenue at the midpoint of its range, and companies on the Databricks AI Index spreading their spend across providers. Each item links to the full piece, which names its sources.

Business Pill · GPUS EXPLAINED

An evergreen Business Engineer explainer on why AI runs on GPUs. It teaches the general idea only and does not discuss the companies in this roundup.

AI Daily Briefing · video

Today’s stories in about a minute (16 Nov 2026).

The key insight: As we read it, Monday’s coverage put prices on AI’s biggest names before their big weeks: Nvidia at 13.3 times its next four quarters of consensus earnings against FactSet’s 18.3 times for the S&P 500, by our arithmetic, and Anthropic at 14.6 times its run-rate revenue at the $74 midpoint once options and stock awards are counted, by our arithmetic. We do not say the threads are linked.

Nvidia, a Day Before Its Report

At $159.92 at 10:24 a.m. ET on 16 November, Nvidia traded at 13.3 times the analyst consensus for its next four quarters of earnings, by our arithmetic, while FactSet puts the S&P 500’s forward 12-month P/E at 18.3. At the index’s multiple, the same $12.02 of consensus earnings would price Nvidia at $219.97 a share, 38% above that trade, by our arithmetic. At Friday’s close the multiple was 13.5 times, against about 22 times for Broadcom and about 40 times for AMD, by our arithmetic. Read: the live multiple and Nvidia against Broadcom and AMD.

TipRanks puts the options market’s expected move for the 17 November results at 8.6% either way; from the $158.84 trade at 10:36 a.m. ET, that range spans 12.1 to 14.4 times the four-quarter consensus, and reaching the index multiple would take a 38.5% rise, 4.5 times the move options imply, by our arithmetic. The earnings bar is $2.49 a share, the consensus of 13 analysts according to Zacks, and Nvidia has reported earnings above that consensus in each of the last four quarters shown on Nasdaq, by 7.4% on average, by our arithmetic. Read: the options range and the $2.49 bar.

At 9:36 a.m. ET, Nasdaq’s last sale of $159.36 was down 2.14% from Friday’s $162.85 close; measured from the 11 November close of $173.73, the last before the 8-K on the chief executive’s resignation, the stock was down 8.27%, or $346.3 billion of market value, by our arithmetic. In Seoul, Samsung Electronics and SK hynix lost about $92.8 billion of market value between them as the KOSPI fell 3.59%, by our arithmetic from Korea Exchange data. Read: the open and the KOSPI.

Broadcom’s market value was 24.8 times its annualized AI semiconductor revenue at the 13 November close, against 11.0 times for Nvidia on its annualized Data Center revenue, by our arithmetic. Broadcom’s AI revenue rose 54% in one quarter, from $10.8 billion to $16.7 billion, its results releases say, while Nvidia’s Data Center revenue rose 18% over its matching quarter, per its 10-Q. Read: the AI-revenue multiples, the quarterly growth and ASICs against GPUs.

Price to the next four quarters of analyst consensus earnings at the 13 November 2026 close: AMD about 40 time
Price to the next four quarters of analyst consensus earnings at the 13 November 2026 close: AMD about 40 times, Broadcom about 22 times, the S&P 500 18.3 times (FactSet) and Nvidia 13.5 times; the company multiples are our arithmetic on Nasdaq consensus forecasts.

Nvidia’s Board, Guarantees and Washington

Nvidia’s board of directors is not considering bringing back Jensen Huang as chief executive, CNBC reported at 17:38 UTC, citing a person familiar with the matter. Earlier, asked about Huang being reinstated, President Donald Trump said “I would certainly not mind seeing that”, according to Reuters. On Kalshi, the contract that pays if Huang is named Nvidia chief executive before 1 January 2029 last traded at 7 cents at 18:06 UTC, half its 14:47 UTC price, by our arithmetic; that is market data, not our forecast. Read: the board report and the president’s remarks.

Nvidia’s latest 10-Q puts the maximum gross exposure under its guarantees at $108.5 billion, including the SB Energy guarantees signed in August. Bank of America analysts estimate that Nvidia “provided $230 billion in financial guarantees”, Reuters reported; that is about 2.1 times the filing’s figure, by our arithmetic, and every partner-facing line in the filing adds up to $189.5 billion, by our sum; Reuters does not give BofA’s method. Read: the guarantee gap.

Summaries of a paywalled report by The Information say Nscale talks cover a guarantee estimated at more than $5 billion that had not been signed by the end of October, more than 1.4 times the $3.5 billion maximum exposure under all of the land, power and shell guarantees for AI clouds in the 10-Q, by our arithmetic. The Financial Times reported, citing two people familiar with its policy, that Nvidia’s Side A directors’ and officers’ cover, up to $900 million, does not include Side B or Side C coverage for the company itself. Read: the Nscale guarantee and the D&O cover.

The Commerce Department’s Bureau of Industry and Security issued an interim final rule requiring a licence to export advanced AI chips to data centres anywhere that give companies headquartered in China, Macau or other Country Group D:5 destinations access to more than 12,000,000 “aggregate TPP” of computing power, and added licence requirements for 5 countries: Indonesia, Malaysia, Singapore, Thailand and Brazil. SemiAnalysis wrote that “more than 1.5GW of Nvidia chips, which are restricted, still ultimately ended up in China”, 3.2 times Epoch AI’s median estimate converted at 700 watts a chip, by our arithmetic. Read: the remote-access rule and the two estimates.

Anthropic’s IPO, Two Days From Pricing

Altimeter Capital plans to invest $5 billion in Anthropic’s initial public offering, Brad Gerstner told CNBC’s “Halftime Report”. At the $74 midpoint, the 825 million shares in the S-1/A come to $61.05 billion, and $5 billion is 8.2% of it, by our arithmetic; an order is not an allocation, and the underwriters decide who gets how many shares. Read: Altimeter’s order.

Anthropic’s 2021 Series A, the $124 million round Jaan Tallinn led, is worth about $35.9 billion at $74 a share, about 288 times the $2.57 per preferred share the round paid, by our arithmetic. At the same midpoint, Anthropic would be valued at 14.6 times its $81 billion run-rate revenue once options and granted stock awards are counted, against 43.7 times for Palantir, by our arithmetic. Read: the Series A and the diluted multiple.

Asked whether OpenAI could end up as the Lyft to Anthropic’s Uber, OpenAI chief executive Sam Altman said at Semafor’s conference: “That is not a worry”. Anthropic’s amended filing shows revenue of $17.0 billion for the three months to 30 September, and OpenAI’s chief financial officer Sarah Friar told Yahoo Finance that third-quarter revenue rose nearly 50% to $10 billion; by our arithmetic Anthropic’s quarter was 1.7 times OpenAI’s. Read: the Semafor exchange.

The trade.xyz pre-IPO contract xyz:ANT had a mark price of $102.06 at 22:05 UTC on 15 November, according to Hyperliquid’s public market data, 30.8% above the $78 top of the filed range, by our arithmetic, while 7 of 10 big 2026 US listings traded below their offer price at the 13 November close, by our count. Separately, Anthropic is cutting its disclosure window for high- and critical-severity bugs from 90 days to 30, for vulnerabilities reported on or after 1 January 2027. Read: the pre-IPO contract, the 2026 IPO class and the disclosure window.

Outages: Microsoft’s Review, Then Chase

Microsoft says the Microsoft 365 outage of 12 and 13 November began with a fix for its network gateway servers that started deploying in July; by our arithmetic that change sat in production for 104 to 134 days, and at the lowest point more than 80% of the gateway’s servers were unresponsive. The Register reported that a final report is due on 20 November. Read: the latent fix and the outage explained.

CyberCube estimates that global cyber insured losses from the outage will most likely be less than $100 million, with a conservative upper bound of $350 million, which is $50 million below the $400 million low end CyberCube gave for the 2024 CrowdStrike outage, by our arithmetic. Read: the insured-loss estimate.

On Monday, a Chase spokesperson said the bank’s services were experiencing a technical bug issue, not a cyber issue, the Associated Press reported, as Downdetector recorded more than 4,700 Chase reports by 2 p.m. ET; nearly 10,000 users had reported an issue with Chase by Monday afternoon, GV Wire reported. Chase made up about 54% of at least 8,700 reports across the four banks FOX 29 listed, by our arithmetic, and no bank has said what the bug was. Read: the Chase outage and the bank’s statement.

What Companies Spend on AI

Anthropic models took 70.1% of spend on proprietary AI models routed through Databricks’ Unity Gateway in the week of 2 November, by our sums of the Databricks AI Index, while OpenAI’s share rose to 27.4% from 10.1% in the week of 20 July. Open-weight models handled 30.9% of tokens sent to Databricks-hosted models, up from 7.3%, and 71.1% of active companies had used models from two or more providers, up from 36%. Read: the spend shares, the open-weight share and the multi-provider shift.

The median Unity Gateway user spent $2.49 in the week of 2 November, up from $0.40 in the week of 20 July, 6.2 times in 15 weeks by our arithmetic. On Ramp, the top 1% of US businesses paying for AI spent a median $10,199.20 per employee per month in October, about 625 times the median firm’s $16.32, by our arithmetic. Read: the median user and the Ramp tiers.

Healthcare AI spend grew 2.5 times to $3.7 billion in 2026, Menlo Ventures estimates, with pharma discovery and research up 28.6 times to $600 million. At the other end, Chegg reported third-quarter revenue of $46.5 million, down 40% from a year earlier and 77.6% below its fourth quarter of 2021, by our arithmetic, and its quarterly report cites ChatGPT and Google’s AI Overview. Read: Menlo’s estimate and Chegg’s filing.

Funds, Filings and Deals

Tiger Global cut its Alphabet Class A shares by 42.3% and its Taiwan Semiconductor ADSs by 35.8% in the third quarter, by our arithmetic from its Form 13F, while Coatue’s 13F shows its Microsoft shares up 119.9%, by our arithmetic, and a new Salesforce position valued at $1.84 billion. HubSpot chief executive Brian Halligan bought 22,985 HubSpot shares, about $5.0 million by our arithmetic, and the Form 4 says the purchase resulted in a short-swing profit under Section 16(b), with $449,641.77 to be paid to HubSpot. Read: Tiger’s 13F, Coatue’s 13F and the HubSpot Form 4.

TensorWave signed a six-year agreement with OpenAI to deliver 457 megawatts of AMD-based capacity, 7.6% of the 6 gigawatts AMD and OpenAI agreed in October 2025, by our arithmetic; no contract value was disclosed. Blackstone Digital Infrastructure Trust agreed to buy 75.1% of joint ventures owning 13 AWS data centres with 647 megawatts, at an aggregate gross value of $5.0 billion, about $7.7 million per megawatt, by our arithmetic, with about half from new mortgage loans. Read: the TensorWave deal and the AWS leaseback.

SpaceX’s 10-Q shows $36,646 million of the $42,474 million purchase consideration for Cursor went to goodwill, 86% by our arithmetic, and puts the Cursor equity awards tied to work after the deal closed at $23,376 million, subject to future forfeitures. Perplexity launched Perplexity Frontier Research with what it calls “considerable computational resources, in excess of 10,000 GPUs”, and health AI startup August raised $12 million from Accel and Peak XV, its founder said. Read: SpaceX’s Cursor accounting, Perplexity’s lab and August’s round.

Courts, the Fed and Security

California Attorney General Rob Bonta sued OpenAI in San Francisco Superior Court; the complaint says OpenAI admitted that it “has lost control of its autonomous AI agents over 80 times during testing” and adds: “It has publicly reported just 27 such incidents to date”. The allegations have not been tested in court. MIT sued Microsoft for patent infringement over a quantum-safe encryption accelerator, its second patent suit against Microsoft this year, by our count of the court archive. Read: California’s suit and MIT’s suit.

Federal Reserve Chair Kevin Warsh said the Fed Board’s staff is modelling how super intelligence could play out in the US economy and described three outcomes; in two of the three, by our count, unemployment rises. Google’s Threat Intelligence Group said threat actors used LLM frameworks “to fully automate vulnerability discovery and exploitation without a human in the loop for the first time”, and that zero-day exploits rose from a 2025 average of 8 a month to 11 a month in 2026, “before spiking to 22 in the month of August”. Read: the Warsh scenarios and the GTIG report.

Slope chart from the Databricks AI Index, week of 20 July vs week of 2 November 2026: companies using two or more AI providers 36% to 71.1%, open-weight share of tokens 7.3% to 30.9%, OpenAI share of proprietary-model spend 10.1% to 27.4%
Three Databricks AI Index measures on different bases, week of 20 July vs week of 2 November 2026: active companies using two or more providers, tokens sent to Databricks-hosted models, and proprietary-model spend (OpenAI’s share is our sum of the index’s model figures).

The Structural Read

As we read it, the Nvidia numbers point two ways at once: the market prices the company below the index on expected earnings, while its own 10-Q gives $108.5 billion of guarantee exposure and Bank of America’s estimate, as Reuters reported it, is $230 billion. Reuters does not give the bank’s method.

As we read it, the Anthropic thread is about who gets the shares and at what price: Altimeter’s planned $5 billion order is 8.2% of the $61.05 billion offering at the midpoint, by our arithmetic, and an order is not an allocation.

As we read it, the Databricks AI Index shows buyers spreading their bets: 71.1% of active companies had used two or more providers by the week of 2 November, up from 36% in the week of 20 July, while Anthropic took 70.1% of proprietary-model spend, by our sums of the index.

Sam Altman, OpenAI chief executive, at Semafor’s Silicon Valley & The World conference, asked about Anthropic’s IPO, as quoted in our piece

“That is not a worry”

Three Implications

NVIDIA HOLDERS Nvidia reports on 17 November against a $2.49 consensus from 13 analysts, per Zacks, and TipRanks puts the options market’s expected move at 8.6% either way.

ANTHROPIC IPO BUYERS Anthropic’s S-1/A offers 825 million shares at $70 to $78; at the midpoint that is $61.05 billion, by our arithmetic, and the underwriters decide who gets how many shares.

ENTERPRISE AI BUYERS On Ramp, the top 1% of US businesses paying for AI spent a median $10,199.20 per employee per month in October, about 625 times the median firm’s $16.32, by our arithmetic.

The Business Engineer Lens

This story maps onto the Business Engineer framework The CFO’s Guide to the Token Economy.

The analysis says that in 2026 AI “became a CFO problem, because the economics finally turned legible”, and defines “an agent is a process that consumes tokens to produce work”, with the job of finance to manage “work produced per token burned”.

As we read it, the day’s data put numbers on that ratio from the buyer’s side: the median Unity Gateway user’s weekly spend rose 6.2 times in 15 weeks, by our arithmetic from the Databricks AI Index, and the top 1% of AI-paying firms on Ramp spent a median $10,199.20 per employee per month. We do not say the analysis predicted these figures. It was published in June 2026 and does not discuss them.

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 threads, and the reading of the day as putting a price on AI’s biggest names, are our grouping, not a finding.

We do not claim that any one story moved prices or caused another. CNBC’s board report rests on one person familiar with the matter; Bank of America’s $230 billion comes without a stated method; the multiples are our arithmetic on consensus forecasts, which can change after the report; Anthropic’s IPO is not yet priced, so its values are our arithmetic at the filed range; the California complaint’s allegations have not been tested in court; and prediction-market and pre-IPO contract prices are market data, not forecasts.

This Week In The AI Supercycle

AI Lost Its Central Banker, the Lab Oligopoly, and the Future of Open Source

This week’s issue: the Nvidia leadership change, read as the AI build-out losing its central banker: who now finances the GPU landlords, backstops the data centres and sponsors open models, and why two labs are becoming the bank.

Read the full issue →

Business Engineer Framework

The CFO’s Guide to the Token Economy

The Business Engineer on why cheaper tokens can still mean bigger AI bills.

Read the analysis →

Get the premium package at 50% off →

The Bottom Line

For anyone pricing the AI build, Monday set two numbers side by side: Nvidia at 13.3 times expected earnings, under the S&P 500’s 18.3 times, and a guarantee book of $108.5 billion in its own filing against a $230 billion bank estimate, a day before the 17 November report.

For the frame behind the spend data, read The CFO’s Guide to the Token Economy on The Business Engineer, on why cheaper tokens can still mean bigger AI bills: read the analysis.

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 15 and 16 November 2026, linked above; every story was built from its own primary source, which it cites and links. We re-read each piece on 16 November 2026 and took every figure here from it. The 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.

Sources: FourWeekMBA: Nvidia’s Forward P/E Is Below the S&P 500’s: 13.3x vs 18.3x · FourWeekMBA: Nvidia at 13.5x Forward Earnings vs Broadcom 22x, AMD 40x · FourWeekMBA: Nvidia’s Options Price an 8.6% Move: 12.1x to 14.4x Earnings · FourWeekMBA: Nvidia Earnings Tuesday: The Bar Is $2.49 After 4 EPS Beats · FourWeekMBA: Nvidia Opens Down 2.1%; $346B Lost Since the 8-K Filing · FourWeekMBA: Samsung, SK Hynix Lose $93B; KOSPI Down 3.6% Before Nvidia · FourWeekMBA: Broadcom Trades at 24.8x Its AI Revenue, Nvidia at 11x · FourWeekMBA: Broadcom’s AI Revenue Grew 54% in a Quarter vs Nvidia’s 18% · FourWeekMBA: Broadcom vs Nvidia Explained: ASICs vs GPUs in 5 Numbers · FourWeekMBA: Nvidia Board Not Weighing Huang Return; Kalshi Price Halves · FourWeekMBA: Trump Open to Huang’s Nvidia Return, 9 Days to Bessent Clock · FourWeekMBA: Nvidia’s 10-Q Shows $108.5B of Guarantees; BofA Says $230B · FourWeekMBA: Nvidia’s $5B+ Nscale Guarantee in Flux, 1.4x Its Cloud Book · FourWeekMBA: Nvidia’s $900M D&O Cover Leaves Out the Company; Cash $56.6B · FourWeekMBA: US Rule Curbs Remote Access to AI Chips, Adds 5 Countries · FourWeekMBA: SemiAnalysis Puts Banned Nvidia Chips in China at 3.2x Epoch · FourWeekMBA: Altimeter’s $5B Anthropic IPO Order Is 8.2% of the Offering · FourWeekMBA: Anthropic’s $124M Series A Is Worth $35.9B at $74, 288x · FourWeekMBA: Anthropic at $74: 14.6x Diluted Run-Rate vs Palantir 43.7x · FourWeekMBA: Altman: Anthropic IPO ‘Not a Worry’; Q3 Sales $17B vs $10B · FourWeekMBA: Anthropic IPO: ANT Perp at $102 Implies $1.34T Valuation · FourWeekMBA: Anthropic IPO Test: 7 of 10 Big 2026 IPOs Trade Below Offer · FourWeekMBA: Anthropic Cuts Bug Disclosure Window to 30 Days From 90 · FourWeekMBA: Microsoft’s July Fix Sat Latent 104-134 Days, Then 80% Froze · FourWeekMBA: Microsoft 365 Outage Explained: Cause, Cost and 7 Numbers · FourWeekMBA: Microsoft Outage Insured Losses Under $100M, CyberCube Says · FourWeekMBA: Chase Outage Reports Near 10,000; JPMorgan Cites Tech Issue · FourWeekMBA: Chase Says Bug, Not Cyberattack, Behind 4,700-Report Outage · FourWeekMBA: Databricks AI Index: Anthropic Has 70% of Closed-Model Spend · FourWeekMBA: Open-Weight AI Hits 31% of Enterprise Tokens; DeepSeek Rises · FourWeekMBA: 71% of Firms Use 2+ AI Model Providers; OpenAI Among Adds · FourWeekMBA: Median AI User’s Weekly Spend Rose 6x Since July: Databricks · FourWeekMBA: Top 1% of Firms Spend $10,199 a Month per Worker on AI · FourWeekMBA: Healthcare AI Spend Hits $3.7B; Drug Discovery Up 28.6x · FourWeekMBA: Chegg Sales Fell 78% From Peak; Filing Cites ChatGPT, Google · FourWeekMBA: Tiger Global Cut Alphabet 42% and TSMC 36% in Q3, 13F Shows · FourWeekMBA: Coatue Grew Microsoft Stake 120% in Q3; New $1.8B Salesforce · FourWeekMBA: HubSpot CEO Buys $5M of Stock, Will Pay HubSpot $450K · FourWeekMBA: OpenAI Signs 457 MW AMD GPU Deal With Neocloud TensorWave · FourWeekMBA: Blackstone’s $5B AWS Leaseback: 647 MW, Half Funded by Debt · FourWeekMBA: SpaceX’s Cursor Bet: 86% Goodwill, $23.4B in Staff Stock · FourWeekMBA: Perplexity Launches Frontier Research With 10,000+ GPUs · FourWeekMBA: August Raises $12M for AI Agent That Spent 405 Hours on Hold · FourWeekMBA: California Sues OpenAI: 80 AI Agent Incidents, 27 Disclosed · FourWeekMBA: MIT’s 2nd 2026 Suit vs Microsoft Targets a Quantum-Safe Chip · FourWeekMBA: Warsh: Fed Models 3 AI Scenarios, 2 With Rising Unemployment · FourWeekMBA: Google Sees First No-Human AI Exploit Tool; Zero-Days Hit 22

Scroll to Top
FourWeekMBA