When the AI Bubble Bursts, Don’t Count on U.S. Consumers

This earnings season has confirmed what we have long suspected: the U.S. consumer appears finally to be buckling under the combined weight of soaring energy prices, depleted savings, and higher everyday‑goods costs.

Even though investors’ attention is focused on artificial intelligence, the U.S. economy remains disproportionately driven by household spending, which accounts for two‑thirds of gross domestic product (GDP). Consequently, if the AI bubble were to burst, the consumers and businesses that represent roughly 90 % of the U.S. economy would be unable to fill the gap.

According to Walmart—still the nation’s largest retailer—the average customer now fuels up with fewer than 10 gallons at a time, a clear signal of strain on household budgets that the company says it has not seen since 2022. Walmart also disclosed that it will have to absorb $175 million in higher fuel costs for its distribution and logistics operations in the first quarter. Costco reported similarly heightened price sensitivity, noting that its members are using its gas stations in record numbers for the first time.

With a decisive resolution to the war in Iran still fragile, inflation is likely to keep climbing; thus, the risks of stagflation—or at least a mild recession—are growing increasingly acute.

These concerns are not evident in official government data. On paper, the economy looks robust. Aggregate consumer spending, unemployment rates, and corporate earnings all appear reassuring. Corporate profits as a share of GDP reached 18.4 % in Q1, the second‑highest level since the 1940s. The S&P 500 continued to set new record highs after each bout of volatility.

However, we have known for some time that much of this strength rests on a narrow base of affluent households and roughly ten large‑cap technology stocks, making both the economy and the equity market overly dependent on a small cohort.

As long as that segment continues to spend the returns from its investments, headline figures will stay solid, largely regardless of what happens to the other 90 % of the population. If that spending were to stop, the fallout could be much more severe.

Nevertheless, this leaves the system with very little resilience should the AI bubble finally burst. The remaining 90 % of mass‑market consumers—those with low‑to‑moderate incomes—are already near the brink. A sizable share of this group is already heavily leveraged, so there is no untapped “new” source of spending, contrary to some narratives.

Personal savings fell to 2.6 % in April, down from 4.9 % a year earlier, the lowest level since 2008, while disposable personal income has also slipped due to higher household utility bills. This puts Federal Reserve Chair Kevin Warsh in a difficult position ahead of his first meeting next week. Futures markets now price a 60 % probability of a rate hike in October. Raising rates to combat inflation—largely driven by an external energy shock—would further dampen consumer spending. Maintaining the status quo, meanwhile, risks allowing inflation to become entrenched.

Most portfolios have over‑weighted U.S. assets, leaving them exposed to a narrow slice of affluent consumers. A more prudent strategy would be to shift capital actively toward diversified global developed markets, where valuations may offer a better cushion. Long‑term outperformance often hinges on incorporating this downside protection through active diversification before broader market shifts take hold.

Standing on the sidelines and hoping the storm will pass will likely leave you caught in the rain.

Matt Rowe, Managing Director, Solutions - Man Group