IMPLICATOR
Torsten Slok, chief economist at Apollo Global Management, published a five-chart note Saturday, October 24, measuring the Magnificent Seven technology stocks against the rest of the S&P 500. The S&P 493 is his label for the index minus Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla.
For those other 493 companies, Slok wrote, the financial payoff from AI was "zero."
His note lands as earnings season opens, with AI spending growing at the top and the rest of the index flat. Gartner estimated more than $3.5 trillion of AI infrastructure spending could be at risk of impairment by 2030 if revenue fails to justify it.
What Changed
- Apollo chief economist Torsten Slok's October 24 note says AI has delivered "zero" financial payoff to S&P 500 companies outside the Magnificent Seven, with revenue per worker and margins flat.
- Ramp data shows AI adoption among businesses on Ramp at 55.9% in September, with two straight months of near-zero growth in token spend.
- A New York Fed analysis finds public-firm real revenue per worker growing 2.2% a year since 2019, but it does not separate the S&P 493 from the Magnificent Seven.
- Elon Musk says SpaceX closed $40 billion of GPU debt and that debt markets have almost entirely shut off firms without substantial revenue outside AI data centers.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
The S&P 493
"Revenue per worker is rising for the Mag 7 but is flat for the S&P 493," Slok wrote. Apollo's first chart, inflation-adjusted to 2019 Q4, puts Magnificent Seven revenue per employee at about $550,000 in 2022, rising to almost $800,000 in 2026, with the S&P 493 around $550,000 since 2019. Profit margins show the same split: rising for the Magnificent Seven since Q3 2023, flat for the rest. "And there are still no signs of AI boosting productivity growth," he added.
"Mainstream corporate America is pulling back and starting to view token spending not as a source of growth but as a cost center that weighs on margins," he noted. Token spending, payments for AI usage billed per token, keeps growing among VC-backed companies. "Mainstream corporate America does not need the latest model and is unwilling to pay for it."
The question, he wrote, is whether, even as models become commoditized, mainstream corporate America can turn increasingly cheap intelligence into higher revenue growth and higher profit margins. "That may be harder than it sounds. AI may be a great calculator, but in business, intelligence is mostly judgment."
FREE · ABOUT FIVE MINUTES
Track where AI spending actually pays off.
Implicator cuts the flood to the developments worth your attention, with concise reporting on what changed and why it matters. Delivered every weekday at 4:45 a.m. Pacific, 7:45 a.m. Eastern.
From San Francisco. No spam. Unsubscribe anytime.
"As long as agentic AI and token spending are confined to a small part of the economy," Slok wrote, "the macroeconomic effects of AI will be very limited."
What the spend data shows
Slok's fourth and fifth charts cite Ramp, whose AI index tracks spend across its business customers. Ara Kharazian, Ramp's lead economist, who previously led economic research at Square, reported on October 14 that AI adoption among businesses on Ramp reached 55.9% in September. "It feels like we're at the top of the S-curve," Kharazian wrote.
Ramp earlier found that 80% of spend on OpenAI and Anthropic comes from 1% of firms. Among the top 1% of AI-spending firms on Ramp, per-employee spend was $9,568 a month in September, down 3.1% from a July high of $9,873. Kharazian recorded two straight months of near-zero growth in token spend and attributes it to price declines at the frontier.
Economy-wide productivity
Jeremy Pearce, a research economist at the New York Fed, published a separate analysis on October 15 using 10-K filings from Compustat. His sample covers manufacturing and services, dropping agriculture, utilities, finance, real estate and public administration. Real revenue per worker grew 0.7% a year from 2010 to 2019 and 2.2% from 2019 to 2025. Nonfarm-business output per hour tracked the same acceleration, from 1.0% to 2.2%. "Across all datasets, the labor productivity uptick is real," Pearce wrote.
Operating margins rose from about 16% of sales in 2019 to 18% in 2025, gross margins from 34% to 37%. Firms with higher R&D intensity saw larger gains. Pearce left open whether the rising margins are "returns to significant investment in intangibles or purely a windfall for firms."
Pearce's sales-weighted sample does not separate the S&P 493 from the Magnificent Seven. His labor-productivity measure is not the same series as the utilization-adjusted total factor productivity Slok charts.
Know someone who'd find this useful? ✉️ Email it to a friend in one click, or they can subscribe free here.
Where the money still flows
On Sunday, October 25, Elon Musk wrote on X that SpaceX had closed $40 billion of debt for GPUs, "but we're investment grade and have Starlink & launch revenue." In a follow-up, he called it "primarily callable debt for GPU purchases that will arrive in Q1 and Q2," adding, "not a major factor this quarter."
Musk named no lenders, rate, maturity or closing date. SpaceX issued no separate announcement. Bloomberg had earlier reported Goldman Sachs leading talks with investors including Apollo Global Management over the financing. The Financial Times cited Apollo as lead investor, with roughly $10 billion of bank loans and $30 billion of investment-grade debt.
Musk was replying to Chamath Palihapitiya, a venture investor writing as an owner of energized data-center land, who called neoclouds "disorganized and poorly capitalized." Neoclouds are newer cloud companies, such as CoreWeave and Lambda, that rent GPU capacity to AI developers, often on borrowed money. "The 200+ neoclouds are going to see guppy-level death and consolidation," Musk wrote. "Some will make it through, but not many." Debt markets, he added, "have almost entirely shut off access to money for any company that doesn't have substantial revenue independent of AI data centers."
When another user on X suggested SpaceX might buy distressed neoclouds, Musk replied: "SpaceX is not buying any neoclouds. Pointless."
SpaceX's AI segment posted $7.56 billion in third-quarter revenue, up from $813 million a year earlier, on total revenue of $14.6 billion. Its CFO guided fourth-quarter capital expenditures of $35 billion to $40 billion.
"The bottom line is that the AI story will be decided in the S&P 493, not the Mag 7, and so far, there is no evidence that revenue growth or margins are improving outside the tech sector," Slok wrote.
Frequently Asked Questions
What is the S&P 493?
It is Torsten Slok's label for the S&P 500 minus the Magnificent Seven: Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. His Apollo note compares the two groups on revenue per worker, profit margins and productivity.
What did Apollo's note find?
Revenue per worker and profit margins are rising for the Magnificent Seven but flat for the S&P 493, and Slok wrote there are still no signs of AI boosting productivity growth. He said mainstream corporate America is starting to treat token spending as a cost center.
Does other data point the other way?
A New York Fed analysis by Jeremy Pearce found real revenue per worker at public firms grew 2.2% a year from 2019 to 2025, and operating margins rose from about 16% to 18% of sales. His sample does not separate the S&P 493 from the Magnificent Seven and uses a different productivity measure than Slok's.
What did Elon Musk say about SpaceX's $40 billion debt?
Musk wrote on X that SpaceX closed $40 billion of debt for GPUs, mostly callable debt for chips arriving in the first and second quarters. He named no lenders, rate, maturity or closing date, and SpaceX issued no separate announcement.
Why does Musk expect neoclouds to struggle?
He wrote that debt markets have almost entirely shut off money for companies without substantial revenue independent of AI data centers, and that the 200-plus neoclouds face "guppy-level death and consolidation," with few surviving.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Free AI briefing · Weekdays
The AI stories that matter, sourced and explained.
Join the Morning Briefing. It goes out every weekday at 4:45 a.m. Pacific, with later sends for the East Coast, Berlin and Tokyo.
Free when you sign up: The Paperclip Compendium, our tested guide to running AI agents.
Free. Unsubscribe in one click.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai
