11/20/2026 | Press release | Distributed by Public on 11/20/2026 02:59
Today's big story: Will 2027 Be A 2022 Repeat For Stocks?
If you own an S&P 500 fund, a big slice of your retirement rides on one bet. The bet is that the hundreds of billions going into AI chips and data centers come back as profit. Here is that story in five lines.
The five biggest AI builders are Microsoft, Meta, Alphabet, Amazon and Oracle. Free cash, used below, is the cash a company keeps after capital spending.
What Do We Actually Know?
Oct 6 was the day after the 10-year Treasury yield peaked at 5.31%. That yield is the interest rate on 10-year US government bonds.
Nvidia has lost 39% of its value since Oct 6. If that one stock holds much of your savings, the drop has hit your own money. In a free 30-minute review, our partner advisory firm can show you ways to limit your loss if the stock falls further and to sell with a smaller tax bill. Book the free review.
The table puts the five lines side by side.
| Line | Now | A year ago |
|---|---|---|
| 1. Builders' profit | $607B | $377B |
| 2. Builders' free cash | $33B | $201B |
| 3. Builders' new debt, 1 year | +$306B | - |
| 4. Anthropic computing due by 2030 | $451.6B | - |
| 5. Nvidia since Oct 6 | -39% | - |
Source: Trefis, prices through Nov 19. Builders are Microsoft, Meta, Alphabet, Amazon and Oracle. Profit is net income and free cash is free cash flow, last four reported quarters vs a year earlier. New debt is borrowed debt, leases excluded, Sep 2025 to Sep 2026 (Oracle to Aug). Anthropic: commitments "expected to be incurred before December 31, 2030", from its filing. Nvidia: Oct 6 to Nov 19 close, dividend-adjusted.
Why Are The Builders Spending So Much?
Because nobody can stop the demand.
Customers want more AI computing than exists. The builders can't wait, because of what happened in cloud, the last big race. There, the companies that built first are the three that now earn about $150 billion a year of operating profit. Operating profit is profit from the business before interest and taxes.
So why does anyone worry? Demand that can't be stopped can still arrive later than the bills.
That happened in 2000. That year Verizon spent $1.11 on its network for every dollar of cash its business brought in. Two years later it cut that spending 31%. The companies selling it the gear fell 89% to 99%. They included Cisco, Juniper and Corning.
Today Amazon's capital spending is $1.28 for every dollar of cash its business brings in. Oracle's is $1.61.
So What Do You Do?
Know which side you own.
When the builders last paused, in 2023, Meta's stock soared. Meta cut its spending 13%. Its free cash more than doubled, and its stock rose 194% that year. When telecom builders paused in 2002, it was the suppliers that collapsed.
So the builders carry the cash risk, and the suppliers carry the pause risk. The builders pay the bills before the profit comes back. The suppliers lose sales if the builders slow down.
If you own an index fund, you own both sides. In our view, that is a reasonable place to be. The risk is being heavy on one side without knowing it. A portfolio full of chip and data-center suppliers is a bet that the spending never slows.
The same logic applies to the AI labs. In our view, own the landlords that rent them computing through the index you already hold. Keep any one lab small enough that a 30% fall in it barely dents your total.
What Would Change The Picture?
Cash. Watch three numbers.
After cash, the next bottleneck is power. We look at it in Where Will AI's Power Come From?
The line to remember: the builders earn more than ever but keep just 5 cents in cash per dollar of profit. In our view, the risk is timing, not demand. Own both sides on purpose, and watch the cash.
Owning both sides on purpose takes a plan you stick with. We have one we are proud of, and we would like to hear how yours compares.
Found A Better Strategy? We'd Love To Learn.
Ours is the Trefis High Quality (HQ) Portfolio: 30 quality stocks in equal weight. Since its September 2020 launch it has returned more than 105%. If yours has done better, tell us about it. If not, come take a look.
For education, not a recommendation to buy or sell.
For questions about this article or underlying research, email [email protected].