|
The Pattern That Predicts Which Side of the Economy Is Lying
At 8:30 a.m. Eastern on Thursday, the Bureau of Economic Analysis released the April PCE report. Headline inflation: 3.8%. Personal saving rate: 2.6%. At 4:15 p.m., Dell Technologies reported $16.1 billion in AI server revenue for the quarter, up 757% year-over-year. The stock jumped 40% after hours.
Two economies, running in opposite directions, sharing the same stock market. It has happened before, and the resolution has always been the same.
The divergence, by the numbers. Corporate profits after tax reached $3.59 trillion at a seasonally adjusted annual rate in the third quarter of 2025, per the BEA. That was near a record. At the same time, the personal saving rate has fallen from 5.5% a year ago to 2.6% in April 2026. Wages grew 3.6% over the past year. Prices grew 3.8%. The gap is small but it runs in the wrong direction: for the first time since 2023, inflation is outrunning paychecks.
The corporate side of the ledger looks nothing like the household side. Dell’s AI server revenue grew 757% year-over-year. Snowflake signed a $6 billion cloud deal in a single announcement. Micron added $150 billion in market cap in one session. These are real businesses generating real cash flow. But the customers who drive two-thirds of GDP are drawing down their savings to pay for groceries and gasoline. American Eagle’s 10% drop on Friday was not a company-specific story. It was a consumer-specific story.
Three times this has happened before. The BEA has tracked corporate profits and household saving since the 1950s. In that span, profits and the saving rate have diverged this sharply only three times.
The first was the late 1990s. Corporate profits surged on the dot-com buildout while the saving rate fell below 4% for the first time since the 1930s. Households were spending freely, leveraging home equity and stock gains. The S&P 500 peaked in March 2000 and lost 49% over the next two and a half years.
The second was 2006 through early 2007. Corporate profits hit a record as a share of GDP. The saving rate fell below 2% and stayed there for eighteen months. Households were extracting cash from rising home values. When the housing market reversed, the consumer cracked first. The S&P peaked in October 2007 and fell 57% by March 2009. The third was late 2019. By September of that year, the University of Michigan Consumer Sentiment Index had fallen to 89.8. S&P 500 earnings were still growing 7%. Lower-income households were already pulling back on discretionary spending before the pandemic provided the trigger. The market fell 34% in five weeks once the shock arrived.
The mechanism is always the same. Corporate profits depend on consumer spending, which depends on the household balance sheet. When the balance sheet is strong, profits and stocks can both rise. When it erodes, a gap opens between what companies are earning today and what they can sustain. The market prices the earnings and ignores the balance sheet. Then something happens: a rate hike, an oil shock, a credit event. The consumer, who had no cushion left, pulls back. Spending contracts, earnings follow, and the gap closes from the top. The current oil shock is exactly the kind of trigger this pattern describes. Gas is above $4.20, and the Iran conflict remains unresolved.
The base case here is not that a crash is imminent. It is that every time this divergence has appeared, the resolution favored the consumer data over the corporate data. Profits are a trailing indicator. The saving rate is a leading one. The household was right in 2000, in 2007, and in 2020. The question is not whether the consumer signal matters. It is how long the market can ignore it.
Today the saving rate is 2.6%, gas is above $4.20 per gallon, and real disposable income is falling. And the S&P 500 just set five consecutive records. My read: the AI buildout is real, the earnings are real, and the consumer data is also real. The market is pricing the first two but not the third. History suggests it eventually will.
A chip company crossed a trillion dollars. A clothing company lost a tenth of its value because shoppers stopped buying jeans. Both happened in the same week, on the same stock exchange. Three times in the past 25 years, that kind of week has marked the start of something the market took months to price in.
|
