Reading a Core PCE Print: The Fed's Actual Inflation Target

Once a month, at 8:30 AM Eastern, the US Bureau of Economic Analysis releases the Personal Income and Outlays report. Buried inside it, on the same line as the personal savings rate and consumer spending, is the number the Federal Reserve actually targets: the Personal Consumption Expenditures price index, and specifically Core PCE.

CPI gets the headlines, the primetime coverage, and the initial tape reaction. But when a Fed governor speaks about the 2 percent inflation target, they are talking about PCE. This article walks the release the same way we walked NFP and CPI: what PCE is, why it differs from CPI, the five numbers that move markets, and how to read the whole release rather than the top-line.

What PCE actually is

The Personal Consumption Expenditures price index measures the change in prices paid for goods and services purchased by, or on behalf of, US consumers. It is produced by the Bureau of Economic Analysis as part of the National Income and Product Accounts. Unlike CPI, which is a stand-alone survey, PCE is derived from the consumer-spending side of the GDP accounts and is reconciled with the rest of the macro framework.

The release comes out on a schedule the BEA publishes each year. It typically lands in the last week of the month, covering the prior month’s data (September’s PCE print reports on August prices). The reference month is the same as the CPI release from two to three weeks earlier, so the market can compare the two prints on the same data point.

Related reading: reading a CPI print for the companion release and the six components that move markets there.

Why the Fed targets PCE, not CPI

This is the single most-misunderstood distinction in retail macro. The Federal Reserve has, since January 2000, targeted PCE inflation rather than CPI. In 2012 the Federal Open Market Committee formally adopted a 2 percent long-run target on total PCE. Four reasons the Fed prefers PCE:

1. Broader coverage. PCE captures spending “by or on behalf of” households. That includes third-party payments the household never sees on a receipt: Medicare, Medicaid, employer-paid health insurance, and other imputed services. CPI covers only out-of-pocket urban consumer spending. Because so much US healthcare spending is third-party, this alone accounts for a meaningful chunk of the CPI-PCE gap.

2. Chain weighting. PCE uses a Fisher-ideal chain index that updates the basket monthly based on actual spending patterns. CPI updates basket weights annually with a lag. When consumers substitute away from expensive goods to cheaper ones (chicken instead of beef), PCE captures the substitution in real time and CPI does not. This makes PCE structurally lower than CPI by roughly 30 to 50 basis points on average.

3. Reconciliation with the rest of the macro framework. PCE is part of the GDP accounts. Nominal spending, real spending, and the PCE deflator all have to add up. That internal consistency is valuable to a central bank that is modeling the whole economy, not just one price index.

4. Revisions. PCE gets revised as new source data arrives, while CPI is rarely revised. For a central bank thinking in terms of trends rather than headline surprises, a revisable series that gets closer to truth over time is preferable to a fixed series that locks in the initial estimate.

The practical result: CPI and PCE both matter, but PCE is the number the Fed is explicitly aiming at. When a Fed member says “we need to see inflation come back to 2 percent,” they mean 2 percent PCE, not 2 percent CPI.

The five numbers that actually matter

1. Headline PCE (all-items)

The all-items PCE price index. Includes food and energy. Consensus is expressed two ways:

Month-over-month percent change (e.g. 0.1%, 0.2%, 0.3%). The short-run signal.
Year-over-year percent change (e.g. 2.4%, 2.6%). The medium-term anchor.

Headline PCE moves markets less than headline CPI does, because CPI got there first. By the time PCE lands, the market has already priced its inflation view for that reference month off the CPI release. PCE surprises are usually smaller in the tape than CPI surprises of the same magnitude, unless PCE tells a materially different story than CPI did.

2. Core PCE (all-items less food and energy)

The Fed’s headline number. Strips food and energy for the same reasons Core CPI does: those components are volatile, supply-driven, and often mean-reverting. Consensus is again expressed MoM and YoY:

Core PCE MoM: 0.1% to 0.3% is the recent range. A 0.4% MoM print is hot and typically dollar-positive, yield-positive, gold-negative.
Core PCE YoY: the number every FOMC statement references. The 2.0% target is a level; the trend (three-month and six-month annualized) is what the committee actually watches month to month.

Core PCE MoM surprises move rates markets even when the print looks small on absolute terms. A 0.3% Core PCE MoM print vs a 0.2% consensus is a genuine hawkish surprise despite the tiny-looking gap. The tenth of a percentage point matters because it compounds: 12 months of 0.3% is roughly 3.7% annualized, well above target; 12 months of 0.2% is roughly 2.4%, close to target.

3. Super-core PCE (core services ex-housing)

The Fed added this decomposition to its lexicon in 2022 and Chair Powell has referenced it in most subsequent press conferences. It measures services prices excluding housing services. Think medical services, transportation services, recreation, financial services, insurance.

Why the Fed cares: super-core is the closest proxy in PCE for wage-driven inflation. It moves slowly, it responds to labor-market slack, and it is the last piece of inflation to come down in a disinflation cycle. If super-core is running hot while headline is cooling, the Fed will lean hawkish because the disinflation is not yet reaching the sticky, wage-sensitive parts of the basket.

Super-core is typically buried in the release tables rather than in the headline summary, but analysts pull it within minutes and it drives the second-wave tape reaction. If you are trading around a PCE print into an FOMC meeting, this is the number to weight highest.

See reading a central bank statement for how the Chair typically frames super-core in the accompanying press conference.

4. Housing services

PCE housing services and CPI shelter measure related but not identical concepts. Both are dominated by rent and owners’ equivalent rent, and both lag the actual rental market by roughly 12 to 18 months because the underlying methodology samples lease renewals rather than new leases.

PCE housing services has a smaller weight than CPI shelter (roughly 15 to 18 percent of PCE vs roughly 33 percent of CPI), so a hot shelter print moves CPI more than a hot housing services print moves PCE. This structural weighting difference is one of the reasons headline CPI and headline PCE frequently print with different signs of surprise even when the underlying inflation story is the same.

Traders who understand this weighting can partly decompose the CPI-PCE gap in real time. A month with a hot CPI print and a cool PCE print is almost always a month where shelter did most of the CPI work; the underlying disinflation trend is closer to what PCE is showing than to what CPI is showing.

5. Personal income, personal spending, and the savings rate

The Personal Income and Outlays report contains more than the PCE deflator. It also publishes:

Personal income (MoM percent change): the aggregate income households received in the month.
Personal spending (MoM percent change): nominal and real consumer spending.
Personal savings rate (percent of disposable income): the residual after spending is subtracted from disposable income.

These three numbers matter for reading the demand side of the economy, not just the price side. A month with hot PCE inflation, weak real spending, and a rising savings rate tells a different story than a month with hot PCE inflation, strong real spending, and a falling savings rate. The first is stagflationary; the second is demand-driven inflation that the Fed has more room to fight.

Rates markets sometimes react to the spending and income data as much as to the price deflators, because the demand story feeds directly into the Fed’s growth-vs-inflation trade-off.

The tape reaction and how it usually unfolds

The 30-second, 5-minute, one-hour framework from the reading an NFP print piece applies to PCE too, with the same important adjustment: PCE lands with more context than CPI does because CPI already primed the tape.

First 30 seconds: algorithmic reaction to headline and Core PCE MoM vs consensus. The initial move is often smaller than a CPI print of the same magnitude of surprise, because the market has partly discounted the PCE result off the CPI data.
First 5 minutes: super-core, housing services, and spending/income get integrated. This is where the initial move often extends or partially reverses.
First hour: the market reconciles PCE with the CPI print from two to three weeks earlier and updates its priced Fed reaction function. Fed-funds futures reprice the next meeting.

A PCE print that confirms the CPI story tends to produce a small, directional continuation. A PCE print that contradicts the CPI story is a bigger event because it forces the market to reassess which release was closer to signal and which was noise.

The Fed reaction function overlay

Every PCE print is read against the Fed’s current stance, exactly as with NFP and CPI:

Fed hawkish (fighting inflation): super-core PCE and Core PCE MoM are the two numbers to weight most. The Fed has explicitly said it needs to see multiple months of sub-0.2% Core PCE MoM prints to be confident inflation is returning to target.
Fed cutting (labor market focus): headline PCE is what gives the political room to cut. A cool headline and a cool Core PCE are what the committee wants to see to justify continued easing.
Fed on hold (data-dependent): PCE is a critical release. Because it is the target, a surprise materially updates the priced path.

The real yields and FX piece walks through the transmission from an inflation surprise to Treasury yields to the dollar. The same mechanics apply to a PCE surprise, with the added weight that PCE is the number the Fed models against directly.

Common misreadings

“PCE is a less important CPI.” No. PCE is the target, and CPI is the earlier, more-frequent leading indicator of what PCE will show. Both matter. CPI moves the tape harder because it publishes first, but PCE is the number the Fed anchors on for its 2 percent goal.

“The CPI-PCE gap is a mistake.” No. The two indexes measure different things using different methods. The gap is structural, and its size and direction contains information. A widening gap tells you something about substitution, healthcare pricing, and shelter methodology. A narrowing gap tells you the two are converging on the same underlying reality.

“Core PCE at 2.5 percent means the Fed is close to done.” Not by itself. The Fed cares about the trend and the composition. A 2.5 percent Core PCE with super-core at 3.5 percent is very different from a 2.5 percent Core PCE with super-core at 2.2 percent. The committee wants to see the wage-driven components come down, not just the headline.

“PCE never surprises.” It surprises less often than CPI, but when it does, the surprise carries more weight per basis point of miss because it is the target. A 10 basis point Core PCE MoM surprise is a genuine event.

“PCE is a lagging release.” True in the sense that it publishes after CPI. But PCE contains information CPI does not: the demand data, the income data, the revisions to prior months. A PCE release with revisions that reshape the trend is not a lagging non-event; it is the release that decides what the trend actually was.

What to do the week of a PCE print

If you are trading around a PCE print:

Read the CPI print from two to three weeks earlier carefully. That is your best estimate of what PCE will show. Analysts publish CPI-to-PCE translation models that back out the likely PCE number from the CPI components.
Know where positioning sits. PCE prints often produce smaller tape moves than CPI prints, so positioning imbalances into a PCE print can create outsized reactions relative to the size of the surprise.
If the PCE print contradicts the CPI print (hot CPI followed by cool PCE, or vice versa), read the internals to figure out which release the Fed is going to weight more. Usually it will be PCE, because that is the target, but the committee will discuss the divergence at the next meeting.
If you are already in a macro position, the same rules from the CPI piece apply. Size ahead of the release, not into the release.

The one-sentence summary

Core PCE is the number the Fed actually targets, it lands two to three weeks after CPI for the same reference month, and the market’s job on release day is to reconcile the PCE print with the CPI read that primed it and update its estimate of the Fed reaction function based on how headline, core, super-core, housing services, and the spending data all land together.

Related reading: reading a CPI print for the companion inflation release two to three weeks earlier, reading an NFP print for the labor-market release the Fed weighs against inflation, reading a central bank statement for the FOMC context every PCE print is read against, and real yields and FX for how a PCE surprise transmits to dollar direction.