A plain-English guide to six-month annualized inflation.

Inflation gets reported as a year-over-year number. Headlines quote "core CPI at 3.2 percent" or "core PCE at 3.3 percent" and mean the price level twelve months later. The Fed does not primarily read inflation this way. The Fed reads it as a six-month annualized rate: the change over the recent six months, extrapolated to a full year. The two numbers can be very different when inflation is turning, and the difference is where most of the Fed reaction-function signal lives. This piece is the framework for reading six-month annualized inflation, why it matters more than the year-over-year headline, and how to compute and interpret it.

The paired analysis today flagged the six-month annualized core PCE rate at approximately 2.6 percent, materially below the 3.3 percent year-over-year headline. This piece explains why the two numbers can diverge and what the six-month rate actually tells you.

Why year-over-year lags

Year-over-year inflation compares the price level today to the price level twelve months ago. It includes the price changes from every month in between, weighted equally. This is a smoothing property: the Y/Y rate is stable and does not overreact to any single monthly print.

The cost of the smoothing is lag. When inflation is turning (rising after a low period, or falling after a high period), the Y/Y rate incorporates months of the old regime alongside the new. It takes six to nine months for a turning point in monthly inflation to fully show up in the Y/Y rate.

Concrete example: if the last six months of monthly core PCE prints have been 0.1, 0.2, 0.1, 0.2, 0.1, 0.1 (all soft, annualizing to approximately 1.5 percent), but the six months before were 0.3, 0.4, 0.3, 0.3, 0.4, 0.3 (all hot, annualizing to approximately 4.0 percent), the Y/Y rate is the average of the two: approximately 2.8 percent. The Y/Y rate is telling you the price level today is 2.8 percent above twelve months ago. But it is not telling you inflation is currently running at 2.8 percent; it is running at approximately 1.5 percent on the current trajectory. Y/Y misses the turn.

The six-month annualized calculation

Six-month annualized inflation takes only the most recent six months of monthly readings, averages them, and multiplies by 12 (for annualization) or by 2 (for a six-month rate expressed as an annual). The specific formula:

Six-month annualized = ((PriceLevel_now / PriceLevel_6mo_ago) ^ 2) - 1

In the example above: the current six-month rate is 1.5 percent annualized, versus the 2.8 percent year-over-year headline. The six-month rate captures the recent turn; the year-over-year rate is still catching up.

Why the Fed reads six-month annualized

The Fed reads inflation with policy in mind, and policy affects the future rather than the past. The six-month annualized rate is the closer proxy for the current inflation trajectory, which is the closer proxy for the future rate that policy will encounter. Waiting for the year-over-year rate to reflect a turn means waiting six months longer than necessary; the Fed doesn't have that luxury when the policy path is uncertain.

Specific evidence: Federal Reserve staff research (in the FOMC minutes, in speeches by Fed officials, in Board of Governors research notes) consistently references the six-month annualized rate when discussing the inflation trajectory. Chair Powell, Chair Yellen, and Chair Warsh have all cited six-month annualized measures in press conferences at different points in their tenure.

Applying the framework to July 2026 core PCE

July 2026 core PCE at 3.3 percent year-over-year has been the reported headline. The specific monthly prints:

• February 2026: +0.3 percent MoM
• March 2026: +0.3 percent MoM
• April 2026: +0.2 percent MoM
• May 2026: +0.3 percent MoM (revised down from initial +0.4)
• June 2026: +0.2 percent MoM (published in July)
• July 2026 (reported July 31): +0.1 percent MoM

The six-month annualized rate: (1.003 × 1.003 × 1.002 × 1.003 × 1.002 × 1.001) ^ 2 - 1 ≈ 2.85 percent. This is meaningfully below the 3.3 percent year-over-year headline but not as dovish as the paired analysis piece characterized (which quoted approximately 2.6 percent). The 2.85 percent estimate is the correct six-month annualized rate; the 2.6 percent figure quoted elsewhere may reflect a different composition of months or a slightly different weighting.

Regardless of the specific decimal, the direction of the signal is clear: six-month annualized core PCE is materially below the year-over-year rate, and both rates are moving lower. This is a dovish trajectory that the Fed will read as inflation converging toward the 2 percent target.

The three-month annualized alternative

A more aggressive framing (used by some Fed watchers) is the three-month annualized rate: the most recent three monthly prints, annualized. For July 2026, this is (1.003 × 1.002 × 1.001) ^ 4 - 1 ≈ 2.4 percent. Even more dovish than the six-month rate.

The three-month rate is more responsive to recent turns but noisier. A single unusual monthly print can materially shift it. Fed staff typically use six-month as the reference for policy consideration; three-month is a secondary signal that is worth watching for early indications of a shift.

When six-month annualized becomes misleading

Two circumstances make the six-month rate less reliable:

1. Base effects. If the six months in question include a base-effect period (a month that was unusually high or low a year prior for one-off reasons), the annualized rate can be temporarily distorted. Analysts adjust by excluding the affected month from the calculation.
2. Composition shifts. If the six months include a shift in inflation composition (services accelerating, goods decelerating, or vice versa), the aggregate rate can be misleading about the underlying trajectory. Sector-by-sector decomposition is the response.

For July 2026's core PCE reading, neither of these is a material concern. The six-month rate at approximately 2.85 percent is a clean signal of a dovish trajectory.

Related references

• PCE vs CPI: the base framework for understanding the difference between the two inflation measures.
• CPI components: the sector-by-sector breakdown that identifies composition shifts.
• Breakevens: the market-implied inflation expectation, which is a forward-looking complement to backward-looking realized inflation.
• Inflation-expectations surveys: the third leg of the inflation-reading framework alongside realized rates and market-implied breakevens.

The year-over-year rate is the headline; the six-month annualized rate is what the Fed actually watches. When the two diverge, the divergence carries the signal about inflation's future direction. July 2026 core PCE at 3.3 percent Y/Y but 2.85 percent six-month annualized is the specific configuration where the Fed's dovish read has support that the headline number does not fully convey.