Core PCE at 3.3%: Will the Fed Raise Rates in September?

The data arrive: inflation persists, while spending stalls

On the evening of August 26, US PCE data for July were released. Headline inflation was 3.7% year on year, slightly above expectations, while core PCE was 3.3%, in line with expectations. Both remained far from the 2% target.

More tellingly, real personal consumption expenditures showed almost no monthly growth. Spending had risen strongly in May and June, but stalled in July. Income increased by 0.4%, the largest rise since January, and the saving rate rose to 3%. It appears the extra income went into savings.

Markets changed direction immediately

After the release, the probability of a September hike jumped from 36% to 42%. The ICE dollar index strengthened to 99.19. Spot gold came under pressure, closing August 26 at 4,594, down 1%, while US equities ended slightly lower.

Two weeks earlier, markets had been talking about fading hike expectations. Now the prospect of another hike was back.

Why is inflation proving difficult to contain?

At 3.3%, core inflation is neither extremely high nor low, but it still cannot reach 2%.

Two new influences are entering the inflation figures.

First, AI: large-scale data-centre construction has raised computer hardware and software prices, supported by Goldman/BEA data. AI demand is finding its way into inflation readings.

Second, share prices: when equities rise, portfolio-management fees rise with them. Those fees increased 21.6% year on year, becoming the second-largest contributor to core PCE inflation.

Inflation is not simply refusing to fall. New influences have entered the last stretch of the journey.

Will the Fed actually raise rates in September?

The 42% figure is market pricing, not a predetermined outcome.

By this morning, CME's September hike probability had fallen back to 36.5%. The 42% figure was the market's immediate response; expectations retreated after investors had time to reconsider.

Interest-rate futures were pricing more tightening by December: 9 basis points by September and 23 basis points over the rest of the year. Markets were therefore reluctant to commit fully to a September hike.

The statistical agency itself also acknowledges problems with the data. On September 30, BEA plans to change PCE methods for computer hardware, portfolio management and legal services. The changes address distortions from AI and equity prices in the current methods. The readings are likely to fall after the revisions.

The more decisive signal will come from the Jackson Hole Economic Policy Symposium, which opens tonight. Markets will focus on whether Warsh keeps further tightening available if the data remain high.

What should we watch next?

1. The Jackson Hole Economic Policy Symposium on Friday.

2. August CPI and payrolls, the final two major data releases before the September meeting.

3. Whether gold can hold above 4,600 after rebounding this morning. If it can, that would suggest markets are less concerned about a rate hike.