What do the data say?
The situation was not significantly altered by yesterday’s PCE inflation report, which showed a rise in both key indicators in line with expectations – headline to 4.1% and core to 3.4%.
Chart 1: PCE Inflation in the United States (2006 - 2026)
Source: XTB Research, 26.06.2026
The trimmed‑mean, a measure of the “trimmed mean”, also rose, albeit much more modestly, to 2.4%. Why is this important? In recent months we have regularly heard from the new Fed Chair that current price‑pressure measurement techniques are far from ideal. As an alternative, he presented the so‑called trimmed mean developed by the Dallas Fed.
“What I’m most interested in is the underlying inflation rate, not the one‑time price change caused by geopolitics or temporary beef price swings.”
The core PCE measure is just a rough scientific guess.
How does it work?
The measure filters out random noise by dynamically and asymmetrically trimming extreme values from the distribution of price changes. Each month the PCE basket is divided into 177 detailed components, ranked in ascending order by their monthly price dynamics and weighted by their expenditure shares. The algorithm then trims the lowest 24% of weights (lower tail) and the highest 31% of weights (upper tail), automatically removing temporary anomalies and month‑specific price spikes. From the remaining “middle” basket, which makes up exactly 45% of the index’s weight mass, a weighted average is calculated – after renormalizing the weights to 100% – which, in annualized form, should show the lasting, long‑term trend of core inflation.
Is this really the best measure?
Institutional and academic analyses find the main advantage of the indicator to be its empirically proven, highest predictive power among core measures for future inflation trends and a stronger, stable correlation with real economic parameters such as labor market tensions. Because the indicator dynamically filters out random shocks without arbitrarily permanently ignoring entire sectors (such as food or energy), it has low volatility and is far less susceptible to later, retrospective revisions of initial readings, making it an attractive tool for designing monetary policy.
The problem arises in conditions of deep structural shocks (e.g., supply shocks, tariff changes). In such an environment, asymmetrically trimming 55% of the basket mass causes the model to largely reject early signals of commodity price increases as anomalous upper tails, thereby systematically underestimating actual inflation and becoming a strongly lagging indicator. Additionally, the high mathematical complexity of this method severely hampers central banks’ clear communication, exposing decision makers to political accusations of intentionally statistically under‑reporting citizens’ real cost of living.
Is the market overestimating the chances of a rate hike?
During his first press conference as FOMC Chair, Warsh repeatedly mentioned that the committee uses “outdated data” and that he remains open to “alternative data sources”, which the market did not pay much attention to. If this message is also emphasized at the next meeting, investors will have a strong argument to temper their expectations for rate hikes. The uncompromising fight to return to the 2% target, assuming the trimmed mean is the reference measure, is now a much easier task.
A fundamental issue is also the Dot Plot itself. Half of the policymakers’ projections currently point to a rate hike before year‑end. However, Kevin Warsh did not place his “dot”, and two hawks – Jeffrey Schmid and Alberto Musalem – have no voting rights in the next two years. The dovish side (Bowman, Jefferson, Cook, etc.) is almost fully represented.
Chart 2: FOMC Dot Plot (2026 - 2028+)
Source: FOMC, 26.06.2026
Moreover, although consumer spending showed solid growth yesterday, it should be remembered that it largely occurs at the expense of rapidly melting savings, and its dynamics in the lowest‑earning third are much lower than what we observe in the main indicators (it does not exceed 1% on an annual basis).
Macro data back in the spotlight?
We have the June HICP inflation from the eurozone (Wednesday) and the U.S. NFP report for the same month. It will be published exceptionally on Thursday, not the usual Friday, due to the 250th anniversary of Independence Day, which extends the weekend for Americans. On Friday (July 3) the U.S. market will be closed, and key overseas readings will be missing.
Chart 3: Change in non‑agricultural vacancies (NFP) and U.S. unemployment rate (2023 - 2026)
Source: XTB Research, 26.06.2026
The latest data do not give us major reasons for concern. Weekly unemployment claims remain low, and ADP and JOLTS data are at multi‑month highs. The only worry might be that after such a strong NFP report for May, it will be hard to match market expectations.
Has the market forgotten about the war?
Perhaps the most important risk that could weigh on further EURUSD decline is a renewed hardening of U.S.–Iran relations. The market currently prices a very optimistic scenario for the conflict’s continuation, which is evident at least in the oil market. We are currently paying just above $72 per Brent barrel, which, given the ongoing risk of a negotiation break by one side and the lack of full reopening of the Strait of Hormuz, seems a risky bet.
If the market is not mistaken in its expectations, the dollar may appear somewhat overvalued. If optimism in that regard is too high, we may soon see a reversal away from risk and a capital outflow to safe havens, supporting the dollar’s strengthening over the past few weeks.
