In partnership with
Welcome back. This edition unpacks why so much of the economy now hinges on AI.
Plus, the math behind OpenAI’s $20 billion revenue gap.
And for paid Macro Talk subscribers: Twilio CMO Chris Koehler on “AI rationalization” plans, managing anxious employees, and Twilio’s strategy for AI agents.
Upgrade your free subscription to unlock the full Macro Talk experience.
The Rising Tide:
Stock Ownership And Being Less Miserable
Traders work on the floor of the New York Stock Exchange at the opening bell on March 12, 2026. (Photo: Timothy A. Clary/AFP via Getty Images)
There’s a “very clear schism” showing up between Americans who own stocks and those who don’t, Bloomberg’s Tracy Alloway wrote this week.
She cites the University of Michigan’s latest consumer sentiment index:
Sentiment among low-wealth consumers — non-stockowners along with those with the smallest stock portfolios — has collapsed about 25% since January.
In contrast, those in the middle and largest tercile of stock holdings have had far more modest declines of less than 10%.
These patterns reflect the fact that strong asset values can cushion the pain of high prices, but only for asset owners.
She also highlights a separate UMich chart, which shows that stockowners are more likely to say they’ll “spend as usual” in response to higher gas prices, while the vast majority of those who don’t own stocks (81%) say they’ll stop or cut back.
How much does this matter
These findings may not be all that surprising, given the role the stock market has played in driving the K-shaped economy.
But because the stock market is driven primarily by the AI trade, this single bet is now propping up two sides of the economy — business investment and consumer spending — Alloway notes.
That’s a lot of concentrated risk, and it all feels even more precarious given mounting questions about whether the biggest AI companies can make their math work.
Speaking of which…
Upgrade your free subscription to unlock the full Macro Talk experience.
A message from our sponsor:
Your SOC 2, handled end to end with AI agents.
Enterprise buyers won’t put your product near their customer data without a SOC 2 report. Twenty-one state privacy laws now hold them accountable for the vendors they share data with, so their diligence lands on you.
Sprinto gets you audit ready in 14 days, across three working sessions. AI agents connect your stack, collect the evidence auditors ask for, and close the gaps as they appear. You approve, they execute.
Sprinto also answers the security questionnaires your buyers send, using the same evidence base, so security reviews stop holding up deals.
No compliance hire. No consultant. Your auditor signs off.
Decoding:
OpenAI’s $20 Billion Revenue Gap
In late September, Axios and other outlets reported that OpenAI was nearing $70 billion in annual recurring revenue (ARR).
Nine days later, on Thursday, the FT reported that OpenAI had recently told investors it was actually closing in on $50 billion in annualized revenue.
The $20 billion difference spooked the market — a reminder of “how nervous investors are that the AI boom is built on a fragile foundation,” Barron’s Adam Clark noted.
So what happened?
If you’re unfamiliar, ARR is a favorite DIY metric for fast-growing startups.
Strictly speaking, ARR is the “total predictable” revenue a company expects to bring in every year from ongoing contracts, per Corporate Finance Institute.
It’s also often used to reflect run rate: one month’s sales (or sometimes a quarter’s) extrapolated over a full year, assuming the pace holds.
“[C]ompanies have a fair amount of latitude” in how they calculate it, Bloomberg’s Sarah Frier and Ryan Vlastelica have pointed out.
That latitude means it’s harder to make apples-to-apples comparisons across companies, and that can lead to fuzzy math.
In this case, Anthropic counts the full value of sales made through partners like Google Cloud, while OpenAI counts only its share, according to Axios.
To compare the two, investors “grossed up” OpenAI’s figure, adding back the partners’ cut, a person familiar with the matter told the FT:
Investors’ efforts to “gross up” OpenAI’s annualised revenue prompted reports that the figure was about $40bn in July, said the person.
The company later told its backers that its annualised revenues had jumped more than 70 per cent since July, prompting reports that the figure was about $70bn at the end of September — a number the company did not deny.
However, the new investor presentation shows close to $30bn annualised revenues in July.
In other words, same 70% growth, but different starting points ($40 billion vs. $30 billion). That’s how you get $70 billion vs. $50 billion.
The larger issue
Both companies need to show there’s enough demand for their models to justify their enormous valuations and spend. That’s why these revenue numbers matter so much.
And as giant corporations like AT&T turn to cheaper open models, investors will be watching closely for any sign that demand is softening.
Speaking of signs…
Strategy Talk:
Twilio CMO — Chris Koehler
Add Twilio — the customer engagement software maker with a $44 billion market cap — to the list of companies taking a harder look at their AI use.
Subscribe to Macro Talk to read the rest.
Become a paying subscriber of Macro Talk to get access to this post and other subscriber-only content.
Upgrade