Everything else here is, in the end, an attempt to guess gross domestic product early. Surveys, benefit claims, lending premiums and the shape of the bond market are all ways of knowing this figure before it exists.

It arrives quarterly, months after the period it describes, and then gets revised more than once. That combination makes it the least useful series on this dashboard for acting on and the most useful one for checking whether anything else was right about the period it covers.

What it actually measures

The comparison is against the same quarter a year earlier, which smooths the seasonal pattern out and builds a full year of memory into every reading. That memory cuts both ways: a weak quarter twelve months ago flatters this one without anything having improved since.

Stagnation is the awkward reading. Growth that has effectively stopped without turning negative produces no headline and describes an economy where nothing is going right. It is the state most likely to be mistaken for a soft landing while it is happening.

Contraction gets confirmed long after it starts. By the time output is measurably shrinking, the leading indicators on this dashboard have usually been signalling for two or three quarters, and markets have generally finished repricing well before the confirmation lands.

The series that arrives last is the one that grades the rest

A number that lands months late and then gets revised is worthless for positioning. It is the standard complaint about it and it is entirely correct.

It is also the reason the number belongs here. Every faster indicator is a guess at this one, and without something to check those guesses against there is no way to learn which of them has actually been worth following.

What it does not tell you

It is revised, sometimes substantially. First estimates have been rewritten by enough to change the story after the fact, so the number that moved markets on the day has more than once turned out not to be the number that stood.

It measures activity, not welfare or distribution. An economy can grow while most participants experience nothing improving, because a rising total says nothing at all about where the increase went.

The recession rule of thumb is not the definition. Two negative quarters is a convention; the formal call weighs depth, breadth and duration and gets made long afterwards.

How to read it

Strong growth. Output is expanding firmly.

Growth. Ordinary expansion.

Stagnation. Growth has effectively stopped.

Recession. Output is shrinking.

GDP Growth sits on the Macro dashboard beside PMI Business Cycle, Labor Market and Inflation.

Common questions

Why is this the worst series for timing?

Because each figure covers three months, lands months later still, and is then rewritten. Anything visible here has usually been traded and forgotten by the time it appears.

Why carry such a slow figure at all?

To settle the argument afterwards. Every quicker indicator here is a guess at this figure, and there is no way to learn which guesses were worth making without the answer to check them against.

Do two negative quarters settle the question?

That is a shorthand people use, not the formal call. The formal call weighs how deep, how widespread and how long the decline was, across several measures, and gets made well after the event.

How much do revisions matter?

Enough to change the story. First estimates have been rewritten substantially, which is another reason to treat this as a record rather than a signal.

What does growth not tell you?

Who got it. Output can expand while most people experience nothing improving, because this measures activity and says nothing at all about how the increase was distributed.