US inflation figures move rapidly through financial headlines, but a percentage without its period, adjustment or source can mislead. This guide explains how the Bureau of Labor Statistics calculates CPI, how to interpret headline and core inflation, and why market reactions cannot be predicted from one number alone.
What the US CPI measures and what it leaves out
The US Consumer Price Index describes price changes for a weighted basket of goods and services purchased by households represented in its urban consumer population. The Bureau of Labor Statistics collects prices from selected outlets and categories and weights them according to consumer spending patterns. It is not a simple arithmetic average of everything displayed in stores, nor is it an exact reading of one family's expenses. A household that drives extensively may experience energy price changes differently from a household that rents in an expensive city. CPI measures price changes over time, not absolute differences in living costs between cities.
The basket includes housing, food, energy, goods and many services. The official tables identify relative importance, so the significance of a dramatic change in one commodity depends on its weight and other components. Gasoline may make a notable contribution to headline inflation, yet it cannot explain the full index without reference to the rest of the basket. BLS measurements are statistical estimates built from samples and specific methods, not a ledger of every consumer transaction. Treat a striking chart on social media as a claim to verify rather than a substitute for the original release.
Read the month-over-month and year-over-year figures correctly
Monthly and annual inflation figures describe different comparison windows. In the official release dated October 14, 2026, for September, the all-items CPI-U increased 0.6% on a seasonally adjusted monthly basis. The twelve-month all-items change was 3.7% before seasonal adjustment. This does not mean prices increased 3.7% during September alone. Nor should a single monthly figure be multiplied by twelve and labelled the actual twelve-month result: it would ignore the sequence of monthly changes and the difference between compounded and simple percentage calculations.
Consider an illustrative index rising from 200 to 201 over one month: the change is 1 divided by 200, or 0.5%. If the index was 194 twelve months earlier, the annual change at 201 would be approximately 3.61%. These are teaching numbers, not the actual September data. When reading a headline, note the reference month, the previous month or year used as a comparison, and whether seasonal adjustments were applied. Accurate figures can still be presented misleadingly when the time horizon or adjustment basis is omitted.
Headline versus core CPI without dismissing household costs
BLS publishes an all-items index as well as an index excluding food and energy. Journalists frequently call the latter core CPI. It is an additional lens intended to reduce the influence of some volatile food and energy movements; it does not imply that groceries and fuel are unimportant to households. The September 2026 statement reported a 0.3% monthly seasonally adjusted increase excluding food and energy, compared with 0.6% for the overall index. Comparing those observations helps identify the contribution of particular sectors but cannot establish the entire macroeconomic trend from a single release.
Inspect shelter, services, goods and the distribution of price changes rather than selecting one spectacular item. A category with a substantial weight can move the index even if several small components hardly change. Likewise, the phrase falling inflation often describes slower price growth rather than falling absolute prices. A move from 5% annual inflation to 3% still means the index has increased over the year. Compare a sequence of releases and inspect the detailed tables before drawing a strong conclusion about the persistence or breadth of price pressure.
Seasonal adjustment, revisions and measurement uncertainty
Travel seasons, holidays, weather and model changeovers create repeated patterns in some prices. BLS publishes unadjusted and seasonally adjusted series where appropriate so that analysts can better interpret consecutive months. The official methodology states that seasonal factors were updated in February 2026 with the January release and that adjusted indexes from the previous five years may be recalculated. If you download a historical series twice, its seasonally adjusted observations may differ because the factors were revised; keep the date and data version alongside any analysis.
CPI is based on sampled prices and cannot mirror every household's shopping basket. A small monthly change should not always support sweeping claims about long-run economic policy, especially when sampling and other uncertainties matter. Analysts should capture the publication date, specific series, the source table and adjustment status, then distinguish observation from inference. The index also cannot explain all dimensions of wellbeing, taxes or income changes. A properly cited official table is preferable to an unlabelled graphic, but even official statistics require careful interpretation.
A practical checklist without turning CPI into a trading signal
First identify the all-items monthly and annual figures. Second compare the food-and-energy-excluded index. Third look for the components and weights most responsible for the change. Only then examine expectations and other economic data. Prices of gold, shares and currencies respond not merely to the number itself but to what market participants anticipated, how other releases develop and prevailing conditions. Sudden moves around announcements can produce high spreads and slippage. There is no rule that a CPI surprise must generate a profitable trade in any particular direction.
The Federal Reserve explains that its longer-run 2% inflation goal is defined in terms of the personal consumption expenditures price index, or PCE, rather than CPI alone. That distinction rules out treating a CPI figure as a mechanical formula for the next interest-rate decision. For research or editorial reporting, keep the release date and official BLS reference with every number and disclose whether an argument reflects an observed fact or an independent interpretation. This guide offers no personalized investment recommendation or prediction of a guaranteed market reaction.
Official resources
No online income or trading system guarantees profits. Results depend on skills, costs and circumstances.
Disclaimer ↗