Core PCE, explained

FRED series

PCEPILFE

PCE price index, less food & energy

Frequency

Monthly

source: BEA, via FRED

Our unit

YoY %

change on the same month a year earlier

Fed target

2.0%

defined on headline PCE, not core

A monthly measure of how much prices are rising for everything US households spend on, minus food and energy. Published by the Bureau of Economic Analysis alongside the broader Personal Income and Outlays report, it is the inflation number the Federal Reserve itself has said it watches most closely when setting interest rates.

“Core” means food and energy prices are stripped out. Both categories swing sharply for reasons that often have little to do with broad-based inflation — a bad harvest, a refinery outage, a war disrupting oil supply — and stripping them out leaves a steadier read on where prices are actually trending underneath the noise.

We take the FRED series PCEPILFE — the Personal Consumption Expenditures Price Index Excluding Food and Energy — and show it as the change on the same month a year earlier rather than the raw index level. That year-over-year line is the default view, drawn over five colour-coded horizontal zones and with an optional Bitcoin price overlay for context.

The chart can also show the same series the other ways FRED publishes it — the index level itself, the change from the month before, and more — and lets you switch between monthly, quarterly, half-yearly and annual. Year-over-year is what virtually everyone means when they cite “Core PCE,” so it is the default view.

Two different government agencies produce two different inflation measures from two different surveys. The Bureau of Labor Statistics publishes CPI from a survey of what households report buying. The Bureau of Economic Analysis builds PCE from data reported by businesses on what is actually sold — a broader base, revised more often.

The practical difference that matters most: PCE accounts for substitution. If beef prices spike and shoppers buy more chicken instead, PCE's weighting shifts to reflect that changed behaviour. CPI uses a more fixed basket and weighting scheme, updated less frequently, which can overstate the pain of a price spike that consumers partly avoided by substituting away from it.

That is why the Fed defines its 2% inflation target on the headline PCE price index rather than CPI — and why Core PCE, the version with food and energy stripped out, is the figure policymakers lean on most for reading the underlying trend.

We draw five horizontal bands behind the line, each keyed to distance from the Fed's target rather than an official policy trigger. At or below 2.09% is green — at or below target, a zone where the Fed is comfortable easing. 2.09% to 2.39% is a second shade of green, still near target. 2.39% to 2.69% is yellow, moderately elevated. 2.69% to 2.99% is orange, still elevated or “sticky.” Above 2.99% is red, high or elevated — the live chart caps its display at 20%, well above anything Core PCE has recorded, so there is plenty of room above that line for the reading to keep climbing.

Core PCE feeds straight into the thing that moves riskier assets: what markets expect the Fed to do next. It is one of the clearest inputs the Fed's rate-setting committee weighs when deciding whether to hold, cut or raise the federal funds rate, and rate decisions shape the money conditions every riskier asset trades in — Bitcoin included.

Cooler readings close to the target raise the market's odds of continued or renewed rate cuts, which historically supports appetite for risk and looser money. Hot, persistent readings raise the odds that tight policy holds for longer, which historically pressures riskier assets. Like any macro relationship, this one shifts the odds rather than settling them — plenty of other forces move price on any given release day.

CPI is the number that makes headlines. Core PCE is the number the Fed actually writes into its policy statements.

Beyond the substitution effect covered above, the two indexes also weight categories differently — PCE gives housing a smaller share of the total than CPI does, and includes some spending categories, like employer-paid health insurance, that CPI does not capture at all. The two measures tend to move in the same direction but rarely show identical numbers, and PCE typically runs a touch cooler than CPI over long stretches.

CPI is released earlier in the month and gets far more headline attention, largely because it lands first. Core PCE is released later, tucked inside the broader Personal Income and Outlays report, and gets less media coverage — despite carrying more weight inside the Fed's own decision-making. We publish both as separate charts because the two measures can pull apart, and reading only the louder one gives an incomplete picture.

Core PCE is a macro context indicator, not a cycle-timing one. It says nothing about where Bitcoin sits in its own market cycle, whether on-chain holders are sitting on profit or loss, or how stretched price is relative to trend — that work is done by the cycle indicators on our Heatmeters.

What Core PCE does is set the policy backdrop those cycle indicators operate in. It is not currently a weighted input to the Bitcoin or Ethereum Heatmeter composite score — it sits alongside CPI, the Fed Funds Rate and the Fed balance sheet in our standalone macro chart set, there for anyone who wants to read the liquidity and policy environment for themselves rather than have it folded into one number.

Others in macro — Interest rates and the supply of money — the backdrop crypto ultimately trades against.

• Unemployment RateLabour-market slack — half of the Fed’s dual mandate.
• Real GDPEconomic growth after inflation, quarter by quarter.
• NFCIA single index of how loose or tight US financial conditions are overall.
• ISM Manufacturing New OrdersForward-looking factory demand, one of the earliest turns in the business cycle.
• Economic Policy UncertaintyHow much uncertainty about policy the press is registering — a proxy for macro risk appetite.

Every indicator we track — the full glossary.