US Earnings Hit 25% Growth Again. That Should Worry You.

The S&P 500 has now posted 25 percent earnings growth for three consecutive quarters. This is the kind of statistic that gets trotted out at investor conferences, cited in morning meetings, and used to justify valuations that would have seemed absurd eighteen months ago. It is also the kind of statistic that demands the kind of scrutiny most people reserve for their mortgage documents.

Let's begin with what we know. US corporate earnings are genuinely expanding. The data is not fabricated. But earnings growth of a quarter, of a sustained quarter, of three sustained quarters at precisely the same rate, suggests something worth examining beyond the headline number. Either American corporations have simultaneously discovered a new model of profitability, or the drivers of that growth are concentrated enough to merit closer inspection.

The easy answer, the one dominating financial media, is artificial intelligence. The AI sector has expanded earnings at rates that would have bankrupted any other industry by sheer gravity alone. Nvidia, Microsoft, and a narrow constellation of technology names have posted growth rates that function as statistical anchors, pulling the broader market average upward through sheer magnitude. This is not hypothetical. The Magnificent Seven, as they have been tediously nicknamed, now represent roughly thirty-two percent of the S&P 500's market capitalization, and their earnings growth rates sit multiples higher than the rest of the market.

Here is where the scrutiny begins. When a handful of companies can move the aggregate earnings number for an index of five hundred firms by that magnitude, the aggregate number stops being a measure of broad-based economic health and starts being a measure of concentration risk. The earnings growth headline becomes true but meaningless, like saying the average temperature in a room is seventy degrees when half the room is an active furnace.

The other five hundred and ninety-three names in the S&P 500 are not posting twenty-five percent earnings growth. They are posting something considerably more modest, something closer to the kind of earnings growth that would be described as healthy in any other era but looks frankly moribund when placed alongside the AI cohort. Financial services are grinding out single-digit growth. Utilities are barely above inflation. The industrial sector is managing low double digits through pricing power and cost discipline, not volume expansion. This is not recession territory, but it is not the broad-based economic expansion that the twenty-five percent headline implies.

The market knows this, or at least parts of it do. The rotation that has defined markets for the past eighteen months—money flowing from everything else into the AI names, then flowing right back out of the AI names on any sign of weakness—reflects a market that understands, at a level deeper than conscious thought, that earnings growth is narrowly sourced. The euphoria is real. So is the fragility.

What happens in quarter four, or quarter one of next year, when the comparison base for AI earnings gets harder? Nvidia posted seventy-five percent earnings growth in the most recent quarter. That growth rate is not sustainable. It is not even historically normal. At some point, even the most transformative technology hits the wall of mature markets and slowing adoption curves. When it does, the earnings growth narrative pivots, and the valuations that were rationalized on the basis of perpetual twenty-five percent expansion start looking expensive.

The market may well handle this transition with grace. AI adoption may continue at rates that sustain extraordinary earnings growth for years. The semiconductor cycle may defy the laws of economic history and climb higher still. But betting on that outcome while ignoring the concentration risk and the mathematical reality of slowing comparables is not analysis. It is faith.

The number is true. The interpretation of that number is what deserves scrutiny.