Here's a thought that might keep you up at night: What if the entire U.S. economy is basically one big bet on artificial intelligence? That's the uncomfortable thesis emerging from Apollo Global Management Inc. (APO), where chief economist Torsten Slok is sounding the alarm about what happens when everyone crowds into the same trade.
His 2026 outlook paints a picture of an economy that has quietly transformed into something unprecedented. The SPDR S&P 500 (SPY) is up 18% this year, which sounds great until you realize it's masking a dangerous concentration. For the first time ever, the 10 largest companies in the S&P 500 make up more than 40% of the index's total market capitalization. And what do these giants have in common? They're all riding the AI wave.
The Concentration Problem
Apollo's concern isn't just about valuation froth. It's about structural dependence. Consider these numbers: Hyperscalers including Microsoft, Google, Meta, Amazon, and Oracle are planning to spend a record 60% of their operating cash flow on AI infrastructure capex. That's not diversification—that's conviction bordering on obsession.
Even more striking, AI-related investment contributed more to U.S. GDP growth than consumer spending in 2025. Think about that for a second. The traditional engine of the American economy—people buying stuff—has been overtaken by companies building data centers and buying chips. AI hasn't just become important; it's become the primary lifeline for economic growth.
When Single Points of Failure Fail
Slok calls this a "single point of failure," and the term fits. If AI demand weakens or those massive capital expenditures fail to deliver immediate productivity gains, Apollo predicts several cascading problems would unfold quickly.
First, there's the market correction scenario. With the S&P 500 so top-heavy, a 20-30% drop in the Magnificent 7 stocks wouldn't just sting—it could erase years of gains across the entire index. When your winners are that large, their stumbles become everyone's problem.
Second, infrastructure spending would freeze almost overnight. Data center construction and chip orders would grind to a halt as companies reassess their AI strategies. The entire supply chain that's been built around this boom would face a sudden stop.
Third, and perhaps most serious, is the recession risk. With non-AI growth already showing weakness, losing AI investment as a growth driver could tip the economy into contraction. Apollo notes that consensus already puts recession probability for 2026 at 30%.
"While we expect the AI cycle to continue in 2026, any rollover would have material negative consequences … We are therefore monitoring all dimensions of the AI story extremely closely," Apollo concluded in their outlook.
The report doesn't predict a bubble pop—it's more of a flashing yellow light. When this much of the economy depends on a single narrative continuing to work, the stakes get uncomfortably high. Whether AI delivers on its promise isn't just a question for tech investors anymore. It's become a question for the entire economy.
