U.S. August Core PCE Falls Short of Expectations: A Cooling Report Created by “Revised Criteria”—Will the Fed Buy It?

According to The Economic Insights, U.S. consumer spending in August experienced the fastest growth in more than a year, while the Federal Reserve’s preferred inflation metric saw a significant drop year-over-year. Data released Wednesday by the U.S. Bureau of Economic Analysis (BEA) show that August personal consumption expenditures (PCE), adjusted for inflation, rose 0.6% month-over-month, the largest single-month gains since March 2025. Nominal PCE increased 0.9%, exceeding market expectations of 0.8%. The Fed’s favored inflation metric, the PCE Price Index, rose 0.3% from the previous month, in line with expectations; while core PCE, excluding food and energy, was up 0.2%, below expectations of 0.3%. The year-on-year declines were even more evident: headline PCE rose 3.4% y/y (forecast: 3.7%), and core PCE rose 3.0% y/y (forecast: 3.3%).

Comprehensive Data: Income Lags Behind Spending, Consumption at Full Steam

From a breakdown, nominal PCE in August increased by $190.8 billion, with goods spending rising by $114.1 billion and services spending rising by $76.7 billion. Automobiles, furniture, and apparel were the main drivers, as consumers continued to spend robustly despite high gasoline prices and increasing overall costs.

Meanwhile, incomes appeared weaker: personal income rose only 0.2% MoM in August (expected 0.4%), with disposable personal income up 0.3%, and real disposable income flat (0.0%) after inflation adjustment. In other words, the acceleration in August consumption was not primarily driven by income growth. Total personal savings was $990.2 billion, with a savings rate of 4.1%.

According to Bank of America credit card data, for the week ended September 19, U.S. consumer spending increased 6.9% YoY, with gasoline spending up 26.5%. Even excluding gasoline, spending still grew by 5.7%—showing consumer resilience with no sign of easing so far.

"Cooling" Factors: How Much Did Three Methodological Adjustments Contribute?

The most noteworthy aspect of this report is that the YoY "decline" coincides with a MoM "rebound." YoY, headline PCE came in at 3.4%, well below the market’s 3.7% expectation, and the July prior figure was sharply revised down from 3.7% to 3.4%. Core PCE YoY came in at 3.0%, under the expected 3.3%, and July was revised from 3.3% to 3.0%—according to FX168’s compilation, this is the lowest core YoY PCE since February.

Yet the MoM is accelerating: headline PCE rose from a revised 0.1% in July to 0.3%, and core PCE rose from 0.1% to 0.2%. The former met expectations, the latter was below the expected 0.3%, but both showed upward movement. This "YoY cooling, MoM heating" divergence results from two sets of revisions introduced simultaneously on Wednesday. First, this data release includes annual revisions to the national accounts covering the past five years (retrospectively from January 2021); second, BEA revised calculation methods for three PCE price index components: portfolio management and investment advisory services, computer software and accessories, and legal services. The largest impact stems from the portfolio management component.

The previous approach directly deflated nominal spending using the industry producer price index, so asset price increases were recorded as “service price rises.” This component saw a YoY surge of up to 21.6% over the past 12 months, becoming the second largest contributor to core PCE inflation. The new methodology uses total hours worked, derived from employment surveys, to measure “real service volume.” UBS economists Alan Detmeister et al. estimate this single change lowered core PCE YoY by about 0.21 percentage points; Goldman Sachs’ Manuel Abecasis et al. estimate that the software and accessories adjustment will lower core PCE YoY by 0.05–0.1 percentage points in May and 0.1–0.2 in December; the legal services component slightly raised YoY by about 0.04 percentage points, partially offsetting the previous two.

In total, this methodological round lowered core PCE YoY by about 0.2–0.3 percentage points—roughly matching the actual drop from 3.3% to 3.0%. In other words, if methodology effects are removed, the inflation reading aligns broadly with expectations; thus, much of the “unexpected cooling” the market saw is statistical rather than fundamental. MarketWatch columnist Jeffry Bartash states it more plainly: the new approach “seems to shave a few tenths off the prior inflation rate, but not enough to suggest a major change in trend”; by any metric, inflation remains too high.

A background worth noting: former Fed Governor Milan noted as early as last December that the current methodology “records as price increases what should be classified as an increase in service quantity.” In May this year, he and Fed economists Barbarino and Diercks systematically reviewed measurement flaws for portable storage devices, video games, and similar categories. However, the timing of this adjustment is particularly sensitive—Trump previously dismissed the Labor Statistics Bureau head over weak employment numbers, and the current president continues to urge rate cuts. Employ America’s Vikas Patel, a progressive think tank member, has called for greater transparency from the BEA in “timing, weighting, and historical revisions.” The question now isn’t just technical validity, but also whether the public trusts statistical independence.

The GDP-side revisions are also significant. Q2 real GDP’s final annualized QoQ figure was revised up by 0.7 percentage points to 2.2% (forecast: 1.5%), with Q1 also revised up to 2.5%. The Q2 PCE price index was revised down to 5.0% (prior: 5.3%), and core PCE to 3.3% (prior: 3.6%). Simultaneous upward revisions to growth and downward to inflation weaken the “stagflation” narrative—right as the Fed resumes rate hikes and needs to assess whether the economy can withstand further tightening.

Market Reaction: Yields Fall, Gold Surges

After the data was released, the reaction was most immediate in the bond markets: the 2-year U.S. Treasury yield fell 4.57 basis points to 4.843%, while the 10-year yield declined about 5 basis points to 5.205%. This pullback is significant—just a day prior (Tuesday), the 30-year Treasury yield hit 5.619% intraday, its highest since June 2002; the 10-year yield at one point approached 5.3%, the highest since 2007, breaking records for the 8th time in September. Tuesday’s yield jump dragged down all three major U.S. stock indices (Dow -0.26%, S&P -0.17%, Nasdaq -0.09%).

Interactive Brokers Senior Economist Jose Torres observed: the stock market “is struggling to hold current levels, but tighter financial conditions are boosting bearish sentiment, while increasing investor interest in downside hedges.”

Gold spiked more than $10 in the short-term, trading at $4,205/oz, and spot gold rose 0.62% to $4,207.75/oz. The dollar index had previously gained 0.17% Tuesday to 101.366, rising 1.5% against a basket of currencies in September; the euro hit a 16-month low against the USD, and after the data was published, the dollar weakened somewhat. On oil, WTI crude fell 3.48% Tuesday to $89.38, Brent lost 2.56% to $102.59 (Middle Eastern supply returned), but rebounded on Wednesday’s Asian session after Trump denied he would ease sanctions on Iran.

Another background indicator not to be ignored: the U.S. September Conference Board Consumer Confidence Index came in at just 81.9 on Tuesday, below expectations and the lowest since 2014—starkly contrasting with robust consumption data, making “how long can consumer spending stay strong” a pressing question.

Policy Outlook: October Rate Hike Expectations Cool

CME FedWatch pricing following the release showed a 65% probability that the Fed will keep rates unchanged on October 28, with a 35% chance of a cumulative 25 basis-point hike, both marking slight declines from pre-release. For the year’s final meeting on December 9, the probability of no change is just 12%, with a 59.5% chance of a 25 bps hike and 28.4% for a cumulative 50 bps hike. Traders have scaled back bets on an October hike.

However, it is more accurate to put this pricing in Tuesday’s context: that day, New York Fed President Williams said, “There’s no need to rush, there is time to collect more information before the October meeting,” and the odds of a rate hike in October had already slipped below 50%. He also stressed that inflation at 3.7% is “undeniably too high,” and that another rate hike this year “may be appropriate”—leaving only the October 28 and December 9 meetings. Evercore analyst Krishna Guha interpreted this as a greater inclination to skip October and opt for a hike in December instead.

Hawks have not relented. Fed Governor Barr stated core PCE met the 2% target in only two of the past 20 months, so “further policy adjustments are likely needed,” attributing some price pressures to AI spending raising chip costs. Governor Cook on September 28 said AI is apparently increasing economic inflationary pressure in the short term, possibly delaying the return to 2%. Cleveland Fed President Harker and Philadelphia Fed President have also expressed concerns that “policy may need to get more restrictive.” The September 16 FOMC voted unanimously (12-0) to raise rates 25 bps to 3.75%–4.00%, with the dot plot showing a year-end median fed funds rate at 3.6%; 16 of 18 decision makers expect at least one more hike by year’s end. The market’s pricing and officials’ divergence set up all the suspense before October 28.

Institutional Perspectives

Morgan Stanley (Michael Feroli): Expecting the tightening cycle to end with just one more hike (in December), “Inflation still appears supply shock-driven, so we do not expect the cycle to extend into next year.” The reason for hiking: “Core PCE inflation has stayed above 3% every month this year, with little recent progress toward the 2% target.” Chair Walsh “has been repeatedly stern on inflation tolerance, which would raise credibility risks for the institution without some action.” Feroli and Abiel Reinhart also note: the savings rate has fallen continuously this year, partly buoyed by the wealth effect from the stock market, but the speed of decline has picked up over the past six months, suggesting some consumers may be straining to keep spending under higher living costs—and if the wealth effect wanes, consumer resilience could be tested.

Citi: Previously expected August core PCE growth of 0.29% MoM and 3.14% YoY, forecasting Q4 2026 core PCE YoY at about 3.1%, below the Fed’s September median forecast of 3.4%.

Nomura: Previously forecast 0.278% MoM and 3.30% YoY, maintaining the view for one more hike in December and no hikes in 2027.

Goldman Sachs: Inflation data may deteriorate somewhat in the coming months before returning to a moderate trend.

Allianz Trade Senior Economist Dan North: Core inflation has yet to show a convincing decline, so the Fed can hardly ignore this level.

RBC Wealth Management: 30-year Treasury yields hitting 6% “is not out of the question.”

A common thread: institutions had warned before the data release that methodological revisions would lower YoY readings, so most regard August’s MoM figure rather than YoY as the real test—at 0.2%, the core MoM sits in a “not hot, not cold” range, insufficient for hawks to back off but also not enough to restore an October hike as consensus.

What’s Next to Watch

Chronologically: The September non-farm payrolls report is due Friday (October 2), with the market expecting about 100,000 new jobs and an unemployment rate rising to 4.2%. This is followed by several weeks of inflation and consumption data; the October 28 FOMC decision; the initial Q3 GDP release on October 29 (coming one day after the rate decision, so more impact for December). In addition, Micron Technology will report earnings after the bell Wednesday—at a time when AI stocks are holding up the Nasdaq and market breadth is at historical lows (the index near 52-week highs while average stocks are far below), this report itself will act as a stress test for the AI trade.

In short: this data does not say “inflation has cooled,” but rather “inflation isn’t hotter than expected and the economy is stronger than anticipated”—giving the Fed justification to wait, and also the foundation for further hikes; the coin for October is still in the air.