US CPI This Week: What It Means for the Fed, US Dollar and Gold

Following the US employment report, market attention is now shifting almost entirely toward inflation. The August NFP provided the Federal Reserve with an important piece of information by showing that the labor market remains more resilient than expected. With 162,000 jobs created compared with only 55,000 expected, unemployment holding at 4.1%, and positive revisions to previous months, labor market weakness is now a less significant obstacle to another rate hike. The question for the Fed has therefore changed. It must now determine whether inflation remains persistent enough to justify further tightening at the September 16 meeting.

This is precisely what makes this week’s data particularly important. The PPI will provide an initial reading of the pressures present at the producer level, before the CPI offers a direct look at the evolution of prices paid by consumers. But this week, looking only at whether the figures come in above or below consensus would be particularly reductive. The real debate within the Fed now concerns the very nature of US inflation, how broadly it is spreading through the economy, and above all whether traditional inflation measures are accurately capturing the underlying trend.

Kevin Warsh clearly brought this debate to the forefront at Jackson Hole. His speech was strongly focused on the persistence of inflation. One of his main arguments is based on the breadth of price increases across the economy. More than half of the roughly 200 categories that make up Core PCE are currently recording inflation above 3%, while around one-quarter are showing increases above 5%. Presented this way, inflation appears much more widespread than the simple monthly evolution of the indexes might suggest.

This interpretation represents an important argument for the more restrictive members of the Fed. If a majority of categories continue to record price increases well above the 2% target, the inflation problem can hardly be described as being entirely concentrated in a few sectors. Under this logic, waiting longer could allow these pressures to remain embedded in the economy, especially now that the labor market has just shown much greater resilience than expected.

However, this measure has an important limitation. Each category is given the same importance regardless of its actual weight in household spending. A category representing several percentage points of the basket can therefore be treated in the same way as another representing an extremely small fraction of expenditures. A multiplication of small categories experiencing high inflation can consequently create the impression of extremely broad inflation while having a much more limited impact on the inflation actually experienced by consumers.

This distinction becomes even more interesting when looking at the source of the recent broadening in price increases. A significant part of it comes from goods, several categories of which carry a relatively small weight in the overall basket. This brings the debate directly back to trade tariffs. Part of the inflation observed this year may be more representative of the effects of a trade shock than of the return of excessively inflationary domestic demand.

This distinction is fundamental for monetary policy. Inflation caused by an overheating economy much more naturally calls for an interest-rate response. Inflation resulting from a temporary change in import costs is more complex to address. The Fed can limit the secondary effects of this increase, but it has much less control over its initial cause. The question then becomes whether this tariff-related pressure spreads durably to other prices or whether it gradually begins to fade.

Recent developments around tariffs reinforce this question. Tariff refunds have increased sharply following the legal changes introduced this year, gradually reducing part of the initial pressure placed on importers. Not all of these savings will obviously be passed directly on to consumers, but the direction is becoming important. If tariffs explained a significant part of the acceleration in goods prices, their influence could gradually diminish over the coming months.

This means that part of the inflation distribution used by Warsh could begin to improve without requiring another significant rate hike. This is precisely one of the factors making the September Fed decision more complex than a simple opposition between high inflation and low inflation.

Services, however, represent a more serious source of concern. The proportion of service categories recording inflation above 3% remains well above the levels observed before the pandemic and shows much less progress. This persistence matters more for the Fed because services inflation depends more heavily on domestic conditions, particularly wages, housing, and domestic demand.

Even this measure, however, must be interpreted carefully. Some of the service prices used in the PCE are not directly observed. Certain categories are imputed from other statistical information, particularly in financial services. They contribute to the official calculation of inflation without necessarily directly representing a price households pay every month.

This issue has become important in the internal Fed debate because Christopher Waller places much less weight on these imputed categories. When these components are removed, underlying inflation based on directly observed prices would currently be around 0.3 percentage point lower than the official Core PCE rate of 3.3%. In July, around half of the monthly increase in Core PCE even came from these imputed categories.

This difference helps explain why two Fed officials can look at the same economy and reach different conclusions. Warsh focuses in particular on the breadth of price increases and sees inflation as still widely embedded across the economy. Waller places greater importance on the recent trend and on directly observed prices, producing a more encouraging picture.

Another measure reinforces this second interpretation: trimmed-mean inflation. This method temporarily removes the categories experiencing the largest increases and the largest declines in order to reduce the influence of extreme movements. It therefore seeks to identify the central trend in inflation rather than giving excessive importance to the most volatile categories.

This measure has slowed significantly in 2026 and now stands much closer to the Federal Reserve’s 2% target than traditional Core PCE. It also has its limitations, but its evolution shows why the diagnosis of US inflation remains much less straightforward than the annual Core PCE rate alone would suggest.

The multi-month inflation trend also provides important information. Core PCE remains elevated at around 3.3% year over year, but its six-month annualized pace has started to slow following two more moderate readings during the summer. Over three months, the improvement appears even more clearly. This means that the annual rate continues to incorporate more inflationary months from the past, while the most recent momentum is beginning to show greater moderation.

This is particularly important for markets. A central bank makes its decisions based on the future direction of inflation rather than looking only at its past level. A 3.3% annual Core PCE rate accompanied by accelerating recent momentum would call for a much more restrictive response than the same 3.3% rate accompanied by a three- and six-month pace gradually converging toward 2%.

Seasonal effects further complicate this interpretation. For several years, US underlying inflation has tended to be stronger during the first part of the year, particularly around January and February, before slowing during the second half. This seasonality should theoretically be corrected in the statistics, but some anomalies appear to persist. Part of the acceleration observed earlier in 2026 could therefore slightly overstate the true underlying inflation trend.

It is within this environment that this week’s PPI and, above all, CPI data must be analyzed. The market is less focused on determining whether inflation still exists than on identifying which of the two interpretations currently present within the Fed best describes the situation. Warsh believes that the breadth of inflationary pressures remains concerning enough to justify a firmer policy stance. Waller places more weight on the recent improvement and considers that, in the absence of another strong inflation report, keeping rates unchanged can still be justified.

The PPI will provide an initial indication of the pressures present upstream in the economy. Another moderate reading would strengthen the idea that the costs faced by businesses are gradually stabilizing. An acceleration would become more concerning if it extends beyond energy categories and shows a more generalized increase in producer prices.

The composition will therefore be essential. An increase driven mainly by oil and energy would have a different meaning from a simultaneous acceleration in goods, services, and the various underlying components. The first scenario could primarily reflect the recent geopolitical shock. The second would indicate inflation that is more broadly embedded in the economy and would provide greater support for Warsh’s interpretation.

This distinction will be even more important with the CPI. The estimates we follow point to a monthly increase in the headline index of around 0.4% in August, with a significant contribution from gasoline prices, which are estimated to have risen by more than 4%. Food prices could also rebound after their slight decline in July.

A high headline CPI reading is therefore already possible without necessarily signaling a major deterioration in underlying inflation. This is precisely why reacting to the headline number alone could be misleading. A 0.4% increase driven primarily by energy does not tell the same story as a 0.4% increase accompanied by a significant acceleration in underlying services and goods.

Core CPI therefore becomes probably the most important figure in the entire report. Estimates point to an increase of around 0.23% for the month, relatively close to July’s pace. On a year-over-year basis, underlying inflation could slow toward 2.4%. If this forecast is confirmed, the report would show headline inflation being disrupted by energy while the fundamental trend continues to improve gradually.

It will then be necessary to look even deeper into the composition. Goods will show whether tariff-related pressures are continuing to pass through to consumers or are beginning to lose importance. Services will help measure the persistence of domestic inflationary pressures. Housing will remain important given its considerable weight in the CPI, while medical care, travel, and accommodation will provide a clearer picture of how broadly inflation is spreading through services.

Wages will also need to be incorporated into this analysis. The latest employment report showed monthly wage growth of 0.3%, while annual wage growth slowed from 3.2% to 3.1%. The labor market therefore surprised strongly to the upside without simultaneously producing an acceleration in wages. For now, this combination limits the argument for a renewed spiral in which an extremely tight labor market pushes wages higher and then drives service prices increasingly upward.

Housing represents another important argument for gradual improvement. Rent growth has slowed, and rental market data suggest that this moderation could continue to feed into the official inflation indexes with a lag. Given the significant weight of housing in inflation measures, this development could continue to place downward pressure on Core CPI over the coming months.

The most hawkish scenario this week would therefore be much more specific than simply a CPI reading above consensus. Ideally, it would require high headline inflation, Core CPI around 0.3% or higher, resilient services inflation, tariff-sensitive categories remaining firm, and sufficiently broad price increases. Such a combination would give greater credibility to Warsh’s interpretation that the inflation problem remains deeply embedded.

Following the 162,000 jobs created in August, this configuration could significantly increase expectations for a September rate hike. The Fed would then have the two elements necessary for another intervention: a labor market strong enough to withstand further restriction and inflation persistent enough to justify that restriction.

The intermediate scenario would be high headline inflation driven mainly by energy, with Core CPI close to 0.2%, moderate services inflation, and continued normalization in the underlying components. This combination would give greater weight to Waller’s interpretation. The Fed would still have the capacity to raise rates, but the economic urgency of a hike would be much less obvious.

In this configuration, a September rate hike could appear more as a preventive decision against inflation risk than as a necessary response to a renewed acceleration already visible in the data. This would also explain why a potential hike could become an isolated move rather than the beginning of a long series of additional tightening measures.

The scenario most supportive of keeping rates unchanged would be another downside surprise in underlying inflation. Core CPI close to 0.2% or lower, accompanied by moderation in services and easing pressure on goods, would significantly strengthen the argument that the recent inflation trend is improving despite the annual rate remaining elevated. The market could then place greater importance on the three- and six-month inflation rates than on the annual figure.

This configuration could produce a particularly significant market reaction because monetary policy expectations remain sensitive following the NFP report. Weaker inflation would show that labor market resilience is not necessarily translating into renewed price pressures. Treasury yields could ease, while the US dollar could lose some of the support coming from rate-hike expectations.

For Gold, this distinction will be fundamental. A broad-based acceleration in inflation accompanied by a significant rise in real yields would represent the most difficult short-term scenario. The market would price in more monetary tightening while US bonds would offer a more attractive real return.

High headline inflation driven mainly by energy could produce a much more complex reaction. If nominal yields rise primarily because inflation expectations increase while real yields react much less, the environment becomes different for Gold. The market could begin to price more heavily the risk that inflation remains difficult to control and that the Fed is forced to keep interest rates higher for longer.

Gold’s reaction after the release will therefore itself provide important information. Strong inflation accompanied by a sustained rise in real yields, an accelerating US dollar, and Gold unable to absorb the pressure would clearly confirm the hawkish interpretation. Strong inflation followed by a limited dollar reaction and Gold recovering quickly would show that the market is also pricing other risks or had already largely incorporated the restrictive scenario.

This week should therefore help distinguish between two interpretations of US inflation. The first, defended by Warsh, considers that the still-significant breadth of price increases shows that the problem remains sufficiently widespread to justify further action. The second considers that this breadth is partly distorted by tariffs, certain low-weight categories, and imputed prices, while more recent measures, trimmed-mean inflation, wages, and rents point to a more significant improvement in the fundamental trend.

The employment report has already answered a first question by showing that the economy has more resilience than expected. The inflation data must now answer the second: does the Fed really need to use that room to raise interest rates as soon as September?

To answer that question correctly, the headline figure will not be enough. The market will need to look at Core CPI, the difference between goods and services, the contribution from energy, tariff-sensitive categories, housing, the breadth of price increases, and above all the three- and six-month trend. It is this composition that will determine whether US inflation is genuinely beginning to deteriorate again or whether still-elevated headline inflation is masking a much more advanced underlying disinflation process.