Reading an economic release

Economic data releases move markets and are reported in a form that strips out most of what makes them interpretable.

The main releases

And what they measure.

Output figures, measuring the total value of goods and services produced, published quarterly in most countries with monthly estimates in some.

Inflation measures, published monthly, with headline and core versions and multiple indices.

Employment data, including unemployment rate, employment level, participation rate and wage growth — each of which can move in different directions simultaneously.

Survey indicators, such as purchasing managers' indices, which are timelier than official data and measure direction rather than level.

Retail sales, industrial production and trade balances.

And central bank communications, which are not data and which frequently matter more.

Expectations, not levels

The key to interpreting the market reaction.

Markets price expectations in advance, which means the reaction depends on the difference between the release and what was expected.

A strong figure that is weaker than expected produces a negative reaction, which appears incoherent unless you know the expectation.

Which means reporting a figure without the consensus expectation conveys little.

And which means the market has already responded before any individual can act on the number.

Revisions

The most under-reported feature.

Initial estimates are based on incomplete data and are revised, sometimes substantially, in subsequent releases.

Output figures in particular are revised repeatedly and can change direction entirely.

Employment figures are revised.

Which means market moves on an initial estimate are frequently reacting to a number that will later be shown to have been wrong.

And it means historical narratives about the economy are frequently based on data that has since changed.

Seasonal adjustment

A technical point that matters.

Most data is seasonally adjusted to remove predictable patterns — retail sales at holidays, employment in agriculture.

The adjustment methods can be disrupted by unusual events, which produced substantial distortions during and after the pandemic.

Comparing a month with the same month a year earlier avoids seasonality and introduces base effects instead.

Which is why annual and monthly rates of change can tell different stories about the same data.

Base effects

A frequent source of misleading headlines.

An annual rate compares with the same period a year earlier, which means an unusual figure a year ago distorts the current rate.

Inflation falling sharply because a large increase a year earlier has dropped out of the comparison is not the same as prices falling.

Which is a distinction that headlines frequently omit and that changes the interpretation entirely.

Headline versus core

For inflation particularly.

Headline includes everything; core typically excludes food and energy, which are volatile and driven by global factors.

Core is used to assess underlying trends and is a poor measure of what households actually experience.

Central banks watch both and also examine services inflation and wage growth as indicators of persistence.

Which means citing a single inflation figure without specifying which conveys less than it appears to.

What matters for an investor

Which is less than the coverage implies.

Individual data releases are noise for a portfolio held over decades.

They are already priced within seconds.

Their initial values are frequently revised.

And the relationship between economic data and market returns is weaker than intuition suggests — markets are forward-looking and respond to changes in expectations rather than to levels.

Which means following releases is an activity rather than an input into decisions.

Where the data is useful

For balance.

Understanding the environment you are living and working in.

Assessing whether wage growth is keeping pace with inflation, which is a household question.

Understanding what central banks are responding to, which affects mortgage and savings rates.

And developing a sense of what normal ranges look like, which makes extreme claims easier to assess.

Reading a release properly

If you do.

Read the statistical office's own summary rather than the news coverage.

Look at the revision to previous periods, not only the new figure.

Look at the trend over several periods rather than one.

Check whether the figure is monthly, quarterly or annualised, since these are frequently confused.

Check the confidence interval, which official statistics generally publish and news coverage generally omits.

And note that a single figure within the margin of error is not a change.

General information only, not investment advice. Investments can fall in value and past performance does not indicate future returns. Consult a regulated financial adviser.