The Federal Reserve's new Chair, Warsh, is attempting to shift the central bank's policy anchor towards a more moderate alternative inflation measure. This significant underlying framework change has sparked market concerns that the Fed might repeat its 2021 mistake of underestimating underlying price pressures.
Current inflation data presents two starkly different faces. The latest figures from the U.S. Commerce Department show the traditional core PCE, excluding volatile food and energy items, rose to 3.3% over the past year, marking its fastest pace since 2023. However, as previously mentioned, the Dallas Fed's trimmed-mean PCE inflation was only 2.3% year-over-year in April.
This statistical divergence directly influences the Fed's interest rate path and financial market expectations for rate cuts. During his April confirmation hearing, Warsh expressed a clear preference for the trimmed-mean PCE, arguing it better filters out one-off shocks like tariffs and geopolitics, thereby supporting a narrative of improving inflation. In contrast, Fed Governor Lisa Cook publicly warned that core inflation indicators are "clearly moving in the wrong direction."
For investors, the current focus is on which inflation thermometer the Fed should trust. If Warsh's preferred policy framework gives greater weight to the trimmed-mean PCE, the logic for the Fed maintaining an accommodative stance or cutting rates in the near term will be strengthened. However, if this gauge currently suffers from systematic bias, this "false sense of security" could lead the Fed to fall behind the inflation curve once again.
The core disagreement pits a cooler alternative measure against traditional inflation worries. While the public widely follows the CPI released by the U.S. Bureau of Labor Statistics due to its early release and linkage to many contracts, Fed policymakers have historically favored the Commerce Department's PCE price index, particularly core PCE. Warsh, however, has called core PCE a "rough swag," arguing it retains too many one-off price distortions.
The Dallas Fed's trimmed-mean PCE, favored by Warsh, attempts to eliminate noise through a systematic filtering mechanism. Unlike core PCE, which permanently excludes food and energy, this measure trims the items with the largest price increases and decreases each month based on the distribution of price changes, retaining the middle portion. Dallas Fed researchers note that in April, trimmed-mean inflation was 0.7 percentage points lower than core PCE, primarily because the measure reduced the weight of goods directly impacted by tariffs.
From Warsh's perspective, current price increases driven by tariff policies, the AI investment boom, and geopolitical shocks are short-term phenomena that should be "looked through" and not warrant policy tightening. The consistently cooling readings of the trimmed-mean PCE provide direct data support for this policy stance.
However, the design of this mechanism harbors the risk of repeating the 2021 error. The Dallas Fed's trimmed-mean PCE had a good forecasting record historically, but its performance during the 2021 inflation surge raised widespread questions. At that time, as inflation soared, the gauge showed a much slower inflation pace than reality, once serving as evidence for policymakers who believed inflation was "transitory."
This mistake stemmed from the indicator's underlying design. Between 1977 and 2009, the magnitude of price declines in the U.S. typically exceeded that of increases. To remove the upward bias caused by this distribution skew, the Dallas Fed designed its measure to trim the top 31% of items with the largest price increases in a given month, but only the bottom 24% with the largest decreases.
However, post-pandemic in 2021, this historical pattern reversed, with the magnitude of price increases beginning to exceed decreases. By mechanically trimming a larger proportion of items with significant price increases, the Dallas Fed's index inadvertently underestimated the actual upward trend in inflation. Now, with a similar divergence reappearing, questions are being raised about whether the gauge is failing again.
Facing the widening data gap once more, research institutions and voices within the Fed have issued warnings. Dallas Fed economist Tyler Atkinson cautioned against excessive optimism from the current trimmed-mean PCE level. He noted that tariffs implemented by the Trump administration have raised prices for a wide range of goods, causing price increases to cover a broader set of items. This may lead the existing trimming rules to exclude too many high-inflation items.
Nomura quantified this bias further in a recent report. The firm pointed out that post-pandemic, core goods prices no longer provide stable deflationary force. Demand for computing and software from the AI investment boom, coupled with increased frequency of corporate price adjustments, makes the distribution of price changes more prone to rightward skew. After adjusting for this bias, the current trimmed-mean inflation is approximately 2.8%, suggesting the official gauge may be underestimating underlying inflation by about 48 basis points.
Data from the left-leaning think tank Employ America also corroborates this potential underestimation. A symmetric trimmed-mean PCE measure constructed by the think tank (trimming equal proportions from the top and bottom of the distribution) reached 3% in April, significantly narrowing the gap with core PCE. Another of its measures, excluding housing and imputed prices, recorded 2.8% in April and has risen year-over-year for 13 consecutive months.
In terms of market impact, Warsh's preference adjustment for inflation indicators represents a framework shift for the Fed in responding to price shocks in the new era. Former Fed economist and head of an inflation research firm, Riccardo Trezzi, stated bluntly that the key question is whether "looking through" price volatility is a principled policy framework or merely a tool to downplay inconvenient inflation data when needed. Trezzi emphasized that evidence inflation has not improved remains strong, as the entire price distribution has shifted upward in recent months.
Other market institutions are also skeptical of the cooling signal from the trimmed-mean PCE. Standard Chartered Bank analysts Steve Englander and Dan Pan argue that, based on historical experience, the gauge is less effective than core PCE at predicting future inflation, making it hard to prove its currently displayed disinflation trend is real. Harvard economist Jason Furman also expressed concern, noting that while referencing alternative measures is not inherently unreasonable, the real risk lies in whether these indicators are being cherry-picked after the fact to fit a particular policy bias.
For financial markets, this shift means the rate-cut narrative has gained a more favorable data point in the short term. If these price shocks are indeed one-off factors, the trimmed-mean PCE would provide justification for the Fed to avoid tightening policy. However, if these alternative measures mask broader demand pressures and structural inflation, over-reliance on them would offer false comfort to markets and could force the Fed into more aggressive tightening measures in the future.
Comments