There is a number that professional investors watch closely, and right now it is telling an uncomfortable story. The S&P 500’s free cash flow yield has compressed to levels not seen since the dot-com era, sitting at roughly 2.58% as of late September 2025 and cited at approximately 2.7% in subsequent Goldman Sachs analysis from August 2026.
Free cash flow yield is essentially how much cash a company generates relative to its stock price. Think of it like a dividend yield, except instead of measuring what gets paid out, it measures what’s actually left over after the business pays its bills and maintains its operations. When that number falls, it means investors are paying more for each dollar of real cash the company produces.
The numbers that should make you pause
Goldman Sachs put the S&P 500 and India’s Nifty 50 on the same slide for a reason: both sit around a 2.7% FCF yield. Europe’s Stoxx 600, by comparison, comes in near 5%. That gap is not a rounding error. It represents a fundamental difference in how much cash investors are getting per dollar invested in equities across different markets.
The more alarming comparison, though, involves bonds. As of late September 2025, the S&P 500’s FCF yield had fallen below the 10-year Treasury yield, which was running around 4.11%. In plain terms, the cash return implied by owning a basket of the largest US companies had dropped below what the government would simply pay you to lend it money for a decade.
The Minneapolis Federal Reserve documented a sharp decline in FCF yields since 2023, tracing much of it to a surge in capital expenditures tied to AI data center construction. The Fed’s framing offered a partial defense of current conditions, noting that today’s yields, while low in isolation, remain consistent with certain long-term postwar averages when viewed across a wider economic lens.
The AI capex machine and its discontents
To understand why yields have compressed this sharply, you have to follow the capital. The largest components of the S&P 500, concentrated in technology, have been spending at extraordinary rates to build out AI infrastructure. These are not small line items. They are multi-hundred-billion-dollar commitments to data centers, custom chips, and energy capacity, all of which hit the capital expenditure column before any revenue from AI products materializes at scale.
This mirrors the pattern that defined the late 1990s. During the dot-com bubble, companies were burning cash to build fiber networks and internet infrastructure. Today’s tech giants are cash-generative businesses that are choosing to spend aggressively on AI, in contrast to dot-com era firms that had low or outright negative FCF yields because they were burning cash just to keep the lights on. The spending still compresses the yield investors receive relative to the price they’re paying.
What this means for how you think about risk
The practical implication for anyone with a portfolio weighted toward large-cap US tech is that the margin for error has narrowed. When an asset yields 5%, a disappointing quarter is painful but absorbable. When it yields 2.7%, prices already embed a lot of optimism about future cash flows, which means any shortfall hits harder.
The geographic divergence Goldman Sachs highlighted adds another layer. European equities at roughly 5% FCF yield are cheaper relative to the cash they generate, a spread that some institutional allocators have already started acting on. With FCF yields below Treasury yields, fixed income has quietly become the higher-yielding option, at least on a current cash return basis.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.

40 minutes ago
6








English (US) ·