Rates Helped. AI Expectations Got Harder.


GOOD MORNING, Wall Street got exactly the macro setup growth stocks wanted Thursday: producer inflation came in softer than expected, Treasury yields fell and September rate-hike odds dropped to roughly one-in-three. The S&P 500 closed at a record and the Nasdaq gained 0.8%. Then Applied Materials reported record revenue, guided well above consensus — and still fell about 5% after hours. The message is becoming clearer: lower rates can help the AI trade, but they cannot lower an expectations bar that has already moved much faster than earnings.


MARKETS | TLDR

  • The S&P 500 closed at a record: The index rose 0.7% to 7,799.57, while the Nasdaq gained 0.8% to 26,803.03 and the Dow added 0.1% to 53,840.14 as softer producer inflation pushed rate expectations lower.
  • Treasury yields fell with Fed-hike odds: The 10-year Treasury yield dropped about 5.6 basis points to 4.636% and the 2-year fell 5.4 basis points to 4.145%; markets cut the implied probability of a September hike to about 34%.
  • Oil finally broke a six-day winning streak: Brent fell more than 2% toward $87 a barrel as weaker demand forecasts and rising U.S. inventories temporarily outweighed persistent Hormuz disruption.

INFLATION RELIEF

PPI Gave the Fed More Room. Markets Used It.

U.S. producer prices were unchanged in July after falling 0.1% in June, better than the 0.2% increase economists expected. Goods prices fell 0.7%, offsetting a 0.2% rise in services, while headline PPI inflation slowed to 4.7% year over year from 5.5% in June. Core PPI also came in slightly softer than expected.

The market reaction was straightforward. Treasury yields fell, rate-sensitive technology stocks rose and the probability of a September Fed hike dropped to roughly 34%. That matters because the equity rally has repeatedly been constrained by the discount rate: when investors expect the Fed to stay on hold, the present value of long-duration earnings rises and richly valued growth stocks get more room.

There is still a caveat. Some PPI components used in the Fed's preferred PCE measure — including portfolio-management services — rose sharply, meaning July core PCE may not be as benign as the headline producer-price number suggests. The PPI report opened a narrower path for the Fed to stay on hold; it did not close the inflation debate.


EXPECTATION BAR

Applied Materials Beat the Quarter. The Stock Still Fell.

Applied Materials reported record fiscal third-quarter revenue of $9.12 billion, up 25% from a year earlier, with non-GAAP EPS of $3.50. The company then guided fourth-quarter revenue to roughly $10.25 billion and adjusted EPS to $4.02 — both above Wall Street expectations. Shares still fell about 5% in extended trading.

That reaction matters because Applied is one of the clearest picks-and-shovels beneficiaries of AI infrastructure spending. Management said rapid AI adoption is driving unprecedented demand for its materials-engineering tools, and it raised its expectations for Semiconductor Systems revenue. Yet the stock had already more than doubled this year, making strong execution less of a surprise and more of a requirement.

The tension is increasingly common across AI. A business can be growing quickly, raising guidance and generating record cash flow while the stock falls because the valuation already assumes an even steeper future trajectory. Lower Treasury yields helped the sector Thursday, but Applied's after-hours move showed that macro relief cannot solve an expectations problem.


HEADLINES

  • Cisco beats, but margins become the story: Cisco reported strong revenue and AI infrastructure demand, but shares fell more than 8% as investors focused on gross-margin compression — another example of AI growth being judged on economics, not just orders.
  • Cerebras shows cloud growth is not the same as chip traction: Cloud revenue nearly quadrupled, but weaker hardware sales pushed shares sharply lower and raised questions about how effectively the company can challenge Nvidia in AI accelerators.
  • SanDisk extends the memory AI trade: Shares jumped more than 15% after the company forecast mid-to-high-teens revenue growth through 2030, reinforcing the idea that AI infrastructure demand is spreading from GPUs into storage and memory.
  • Oil relief remains conditional: Brent fell as U.S. inventories rose and demand forecasts weakened, but Hormuz traffic remained heavily disrupted and Iran continued to claim control over the strait.
  • Big Tech capex is still the anchor: Reuters estimated spending by major technology companies is set to exceed $740 billion this year, keeping infrastructure demand strong while simultaneously raising the bar for every supplier tied to the buildout.

UPCOMING

  • Retail sales — August 14: July spending data will test whether the consumer is still resilient enough to support growth after softer inflation gave the Fed more room to stay on hold.
  • University of Michigan sentiment — August 14: Consumer confidence and inflation expectations will show whether households are feeling relief from softer goods prices or pressure from higher energy costs.
  • Home Depot earnings — August 18: The retailer will provide a rate-sensitive read on housing turnover and large renovation projects as long-term borrowing costs remain elevated.
  • Fed minutes — August 19: July meeting minutes will help show whether policymakers were already leaning toward another hike before this week's softer CPI and PPI data.

DEEP INSIGTHS

Producer Price Index — July 2026

The primary BLS release is useful because the headline was softer than expected while the internal composition was mixed: goods prices fell, services rose and some components relevant to core PCE remained firm. That explains why markets cut September hike odds without declaring inflation solved.

Applied Materials Q3 FY2026 Results

Read the release for the gap between operating performance and market expectations. Revenue hit a record, EPS rose 41% on a non-GAAP basis and Q4 guidance came in well above consensus — yet the stock sold off after hours. It is a clean case study of how difficult the AI expectation bar has become.

background

Subscribe to Markets in Minutes