AnalysisIndicators8 min read

Reading inflation data: CPI, PPI, and PCE

Three reports, three methodologies, one decision. Which one the Fed actually targets, why core matters more than headline, and the components to watch when the topline lies.

Published May 28, 2026

TL;DR

CPI is the inflation report markets trade. PPI is the leading indicator. PCE is what the Fed actually targets. Each is computed differently, weighted differently, and tells you something the others don't. Reading them in isolation is how traders get whipsawed. Reading them as a system — leading indicator → consumer prices → Fed-target — is how you anticipate the next policy move.

Three reports, three jobs

CPI (Consumer Price Index) measures the price of a fixed basket of goods and services bought by urban consumers. PPI (Producer Price Index) measures wholesale prices charged by domestic producers. PCE (Personal Consumption Expenditures) measures consumer spending but with a flexible basket that adjusts as substitution behaviour changes. The Fed's 2% inflation target is the PCE number, not CPI. This single fact dictates which report actually moves policy.

Why core matters more than headline

Each report has a 'headline' version (everything) and a 'core' version (excluding food and energy). Food and energy are volatile and supply-driven — they move on weather, geopolitics, and OPEC decisions, not on demand. The Fed doesn't want to overreact to a hurricane in the Gulf of Mexico or an oil-supply shock, so it focuses on core. Markets follow. When you see headline CPI come in hot but core soft, the policy implication is muted.

→Headline = all components, including food and energy.
→Core = headline minus food and energy. The Fed's preferred reading.
→Supercore = core services minus housing. Powell's frequently-cited 'sticky inflation' measure.

The PPI lead

PPI typically leads CPI by 1-3 months because producer-price changes have to work through inventory, contracts, and pricing decisions before they hit consumer shelves. A surprise jump in PPI core is one of the earliest warning signs that next month's CPI will be hotter than expected. The relationship has weakened since the pandemic-era supply shocks but still works directionally.

The components that move policy

Inside CPI, three components carry outsized weight in the Fed's reaction function. Shelter (rent and owners' equivalent rent) is roughly a third of the index and lags real-time market rents by 12-18 months — meaning even a cooling housing market can keep CPI hot for months. Wages and core services ex-housing are the supercore measure Powell tracks because they're the most demand-sensitive and the least supply-driven. Goods inflation matters less in a service-economy print but matters more during supply shocks.

Reading the release calendar

CPI prints monthly, around the 10th-15th, before the FOMC meeting in months that have one. The Fed sees the CPI report as part of its inputs but doesn't react to a single month — the question they ask is whether the three-month annualised trend has shifted. PCE prints two weeks after CPI, so it adds little new information to a market that already traded CPI hard. The actionable trade is around CPI; the confirmation trade is around PCE.

Worked example

Walk through a hot CPI print

CPI comes in at +0.4% MoM (consensus +0.2%), with headline year-over-year at 3.2% (consensus 3.0%). Core CPI is +0.3% MoM (consensus +0.2%). What does each line tell you, and which one matters most?

1Headline beat+0.4% vs +0.2% — twice the expected pace, but largely energy-driven (gasoline +5%)
2Core beat+0.3% vs +0.2% — smaller, but the one the Fed weighs
3Component checkShelter cooling (+0.2% vs +0.4% prior); supercore reaccelerating
4Market reaction (T+0)2-year yield +10bp; SPX −0.7%; DXY +0.4%; gold −0.5%
5Component-driven reversalBy midday, focus shifts to the cooling shelter print → SPX recovers half the loss
6Fed implicationPCE in two weeks now the next checkpoint — if it confirms core, hike probability rises ~10ppts

Takeaway

The headline drove the open. The components drove the close. Traders who anchored on the −0.7% SPX move missed the cooling-shelter read that flipped the narrative within hours.

Common mistakes

What to avoid

!Trading the headline before reading the components. Energy can drag headline either way without changing the Fed's view.
!Treating CPI as the Fed's target. The 2% target is PCE — they're correlated but not identical.
!Ignoring shelter's lag. Housing market cooling doesn't show up in CPI shelter for 12-18 months.
!Conflating goods and services inflation. A goods-driven print signals supply chain dynamics; a services-driven print signals demand.
!Reacting to a single month. The Fed looks at 3- and 6-month annualised trends — single-month surprises usually mean-revert.

Self-check

Test yourself

Q1If CPI prints hot but the surprise is entirely in energy, why does the market reaction often fade?+

Because the Fed focuses on core (ex-food, ex-energy) and on supercore (core services ex-housing). An energy-driven beat doesn't change the policy path, so the initial 'hot CPI' move reverses as traders re-read the components.

Q2Why does PCE matter even though CPI prints first and gets all the market attention?+

PCE is the inflation measure in the Fed's 2% target. CPI is the trade; PCE is the confirmation. The Fed's reaction function is built on PCE trends, even though markets trade the CPI release because it lands earlier.

Q3What does it mean if shelter inflation is rising sharply but real-time rent indices have been flat for six months?+

It means the CPI shelter component is still working through its 12-18 month lag. Real-time rent changes show up in CPI a year or more later. A cooling rental market today implies cooling CPI shelter only in 2025-2026.

Q4Why is PPI considered a leading indicator for CPI?+

Producer prices change first because they reflect input costs that producers eventually pass through to consumers via shelf prices, contracts, and inventory turnover. PPI typically leads CPI by 1-3 months.

Keep reading