The Week AI Spending Stopped Being a Free Pass

Markets spent the back half of summer torn between two very different stories. On one hand, company earnings have been outstanding. Nearly 9 out of 10 S&P 500 companies beat their earnings estimates, growth came in above 35% versus an expected 24%, and profit margins kept rising. Consumer spending stayed healthy across all income levels. Despite all this good news, the market stopped rewarding companies the way it used to.

The turning point was one large platform company’s earnings. The results were genuinely impressive and its backlog of future business jumped several-fold in a year, showing demand still bigger than what it can supply. But instead of celebrating this, investors focused on the fact that it is now burning cash and spending heavily on AI infrastructure, and the stock got punished.

Investors are no longer satisfied with big spending alone. They want to see proof that this spending is actually paying off. Even the bond market picked up on this shift, demanding higher returns to lend to big tech companies.

There is a bigger question sitting underneath all of this. AI intelligence itself is becoming cheaper and more widely available. Open-source models are becoming serious competitors. This has caused money to rotate away from AI-heavy growth stocks and into small caps, value stocks, industrials, and defensive sectors.

Global events added more pressure. Rising tensions around Iran pushed oil prices toward the high $80s, keeping inflation worries alive and raising the odds of a surprise Fed rate hike right before the midterm elections.