US K-shaped Economy: An Attractive Divide... But Less Clear Than It Seems

An Economy of Paradoxes Rather Than Disruptions

The contrast is striking: on one side, stock markets are near their record highs, boosted by AI and rising profits; on the other, household sentiment, measured notably by the University of Michigan index, remains surprisingly low. This gap fuels the idea of a « K-shaped » economy, divided between the winners of innovation and the losers of real life.

However, this dramatic theory doesn't hold up well when examining the numbers. While weak consumption signals suggest caution, they do not confirm a clear break. Retail sales grew by only +0.2% in September (including inflation), and iconic companies like Home Depot and Chipotle warn about the weakened purchasing power of low-income households. To see this as a structural downturn is a big leap.

Real-time consumption data, provided by some private companies, significantly temper this narrative. The company Numerator notes that household spending for those earning over $100,000 per year increased by +4.3% in Q3 2025, compared to +3.8% for those earning under $60,000. There is a gap, but it's modest. Even the sentiment expressed by consumers varies less by income and more by measurement methodologies: wealthy households demonstrate distrust comparable to that of lower-income ones.

Since 2022, behaviors have been relatively stable: low confidence, slightly slowing but still strong consumption. After a growth of +2.9% in 2024 excluding inflation, it's expected to increase by +2.5% in 2025, despite tariff shocks, and consensus projects +1.9% in 2026. Recent revisions even suggest a slightly stronger scenario than anticipated.

At the heart of this apparent inconsistency, multiple realities overlap. Real estate, in particular, affects household perceptions: despite the rate hikes of 2022-2023, prices have not corrected. The US is experiencing an affordability crisis similar to the 1980s when long-term rates exceeded 10%. The rise in costs—housing prices, insurance, mandatory expenses—fuels a discomfort that's more psychological than macroeconomic.

A Convenient Narrative for the Markets, Less So for Economic Analysis

The strength of the « K-economy » lies more in its narrative power than in its statistical precision. It reflects an economy where conflicting signals abound: high profits, inflation contained in some areas, tariff policies akin to a tax increase, a less strained labor market, and the enduring impact of the COVID-induced inflation trauma on household sentiment.

The historical divergence between the two main confidence indicators is symptomatic of this: the University of Michigan's indicator places central emphasis on inflation sentiment, while the Conference Board's indicator gives a more balanced view of income, labor market, and prospects. Since the pandemic, their paths have only continued to diverge, resulting in a very deteriorated perception, even when the fundamentals are not worsening at the same rate.

For financial markets, this disassociation is far from being bad news. Weak confidence strengthens the conviction that the Federal Reserve will remain extremely responsive to any signs of a slowdown. A slightly disappointing retail sales number was enough to reignite expectations for a rate cut at the December 10 meeting. In an environment where corporate profits are strong, this responsiveness acts as a stabilizer for risky assets.

The resilience of the markets is also tied to the renewed role of government bonds. They have become an effective hedge against a potentially more pronounced slowdown in the US economy. In the context of high valuations on equity assets, this mechanism reinforces the sentiment that the Fed will not allow financial conditions to tighten to the point of breaking momentum.

Should we then definitively bury the « K-economy » theory? Not quite. It highlights some truths about the structural vulnerabilities of the US economy: dependence on AI for productivity growth, the redistributive effect of financial markets, the fragility of real purchasing power, and the increased selectivity of consumers. However, it overestimates the speed of divergence among different household categories and overlooks the overall coherence of an economic model still driven by consumption.

Ultimately, rather than a « K » economy, the data describes an American economy in a state of permanent paradox: households are worried but continue to spend, the labor market is solid but less dynamic, real estate is stagnant yet resilient, and the Fed is vigilant yet constrained. It's an unstable configuration, but far from a clear-cut fracture.