Inflation and unemployment releases are among the few scheduled events that reliably move entire markets. They are also routinely misinterpreted, partly because they are summarised as single numbers when they are actually large collections of estimates, and partly because the headline figure is often the least informative part.
Both figures are estimates built from samples, shaped by methodological choices, and revised after publication. That does not make them unreliable, but it does mean the sensible unit of analysis is the trend across several releases rather than any single print.
What a consumer price index actually is
A consumer price index tracks the cost of a fixed basket of goods and services intended to represent typical household spending. Statistical agencies collect a large number of prices, weight each category by how much of household spending it represents, and report the change over time.
Several consequences follow from that construction, and they explain most of the confusion around inflation data.
- It is an average across a population, not a description of your costs. If your spending is unusually concentrated in a category, your personal inflation rate can differ substantially from the published one, and neither figure is wrong.
- Weights are updated periodically, not continuously, so a basket can lag genuine shifts in how people spend.
- Housing is treated very differently between countries, and in several major economies including the United States it is the largest single component while also being measured in ways that respond slowly, since owner-occupied costs are estimated using rental equivalents. The euro area index excludes owner-occupied housing altogether. So a housing-driven inflation story means different things depending on whose index you are reading.
- Quality adjustments are applied when a product improves, which is methodologically sound and also a common source of public disagreement with the figures.
Because of these features, month-to-month movements contain a good deal of noise, and single-month surprises are frequently reversed.
Headline, core, and why both exist
Headline inflation includes everything in the basket. Core inflation typically excludes food and energy. This is often described as excluding the things people actually buy, which is a fair complaint and a misunderstanding of the purpose.
Food and energy prices are volatile and driven substantially by global supply conditions, weather, and geopolitics rather than by domestic demand. A central bank raising interest rates cannot influence a drought or a shipping disruption. Core inflation is an attempt to see the part of price pressure that monetary policy can actually act on, and that is more likely to persist.
The practical implication is to read both and note whether they agree. Headline above core usually means an external supply shock, which central banks often look through unless it lasts long enough to change expectations. Core above headline means underlying domestic pressure, which is what actually prompts policy response. A gap that closes over several months tells you the shock is passing through and fading.
Two further refinements are worth knowing. Services inflation tends to be stickier than goods inflation, because services costs are dominated by wages, so a decline led entirely by goods may be less durable than it appears. And the annual rate depends on what happened twelve months ago as much as on the current month, which means base effects can drive a rate up or down without any change in current momentum.
What an unemployment rate leaves out
The unemployment rate is the share of the labour force that is without work and actively seeking it. The definition of the labour force is what makes the number subtle, because people who are not looking for work are excluded from the denominator entirely.
This produces a well-known and genuinely counterintuitive effect. If discouraged workers stop searching, they leave the labour force, and the unemployment rate can fall while the number of people working has not increased at all. Conversely, when conditions improve and people re-enter the search, the rate can rise for a good reason.
For this reason the participation rate, which measures the share of the working-age population in the labour force, should be read alongside the unemployment rate. A falling unemployment rate accompanied by a falling participation rate is a weaker signal than the headline suggests.
Two more caveats matter. The rate does not capture underemployment, meaning people working fewer hours than they want or in roles below their qualifications, which is why some agencies publish broader measures. And labour market data is famously subject to revision, sometimes substantially, so a single print should be treated as provisional.
Why markets react to the surprise, not the number
An inflation rate of 3 percent is neither good nor bad news in itself. If the market expected 3 percent, the release contains no information and prices should not move. If the market expected 2.5 percent, the same 3 percent figure is a meaningful upside surprise.
This is why financial reporting always frames releases against expectations, and why you cannot interpret a market reaction without knowing the consensus forecast. It also explains reactions that appear backwards, such as equities rising on weak employment data, which happens when the market reads weakness as reducing the pressure for further rate increases. The data was bad for the economy and, in the market's reading, good for the discount rate.
That relationship is not stable, and assuming it is has cost people money. Weak employment data supports equities when the dominant market concern is interest rates. The same data hurts equities when the dominant concern is recession and corporate earnings. Which regime you are in determines the sign of the reaction, and regimes change without announcement.
A workable way to read a release
- Find the consensus expectation before the release, so you can identify the surprise rather than react to the level.
- Compare headline and core, and note which is higher and whether the gap is widening or closing.
- Look at the trend across several months instead of the single print, because one month is mostly noise.
- Check the composition. Which categories drove the change, and are those categories volatile or persistent?
- For labour data, read participation alongside the unemployment rate, and treat the figures as provisional given the revision history.
- Ask what this does to the expected policy path, since that is the channel through which the data reaches your holdings.
The purpose of this discipline is mainly to prevent unnecessary action. Most single data releases do not change anything about a long-term portfolio, and the ones that do tend to be part of a trend visible across several months rather than a surprise in one print.
General educational information, not investment advice. Statistical methodologies, basket weights, and definitions differ between countries and are revised over time.