US GDP and recession calls
GDP is the number everyone quotes and almost nobody reads correctly - starting with the fact that the US publishes it three times per quarter and annualises it while most of the world does not.
At a glance
- Published by
- US Bureau of Economic Analysis, quarterly
- Released
- Three estimates per quarter, roughly a month apart
- Quoted as
- Real GDP, annualised quarter-over-quarter percent change
- Recessions
- Dated by the NBER, not by a rule about negative quarters
What the number actually measures
Gross domestic product is the value of everything produced in the country over a period. Real GDP strips out price changes so that what remains is a change in volume rather than in cost - which is why a nominal figure and a real one can move in opposite directions during a burst of inflation.
The US quotes it annualised: a quarter in which output rose half a percent is reported as roughly two percent, because that is what four such quarters would compound to. Most other countries report the plain quarterly change. Comparing a US headline with a European one without converting is one of the most common errors in macro commentary.
There is a companion measure worth knowing. Gross domestic income adds up the income side of the same activity and should, in principle, equal GDP. It does not, and the gap between them has at times been large enough that the two told different stories about whether the economy was growing.
A US GDP headline is annualised; most other countries' are not. Divide by roughly four before comparing.
Who acts on it
GDP is the reference for whether the economy is expanding, which makes it an input to monetary policy, to fiscal planning, and to every debt ratio expressed as a share of output. It is also the least timely of the major indicators: the first estimate of a quarter arrives a month after that quarter ended, by which point markets have already read a dozen monthly series that fed into it.
That lag is why GDP moves markets less than payrolls or CPI despite being the broader measure. By release day, most of its content has been anticipated - which also makes the composition, rather than the headline, the interesting part.
- Consumer spendingServices roughly two thirds of this
- 66%
- Private investmentBusiness fixed, housing, inventories
- 17.5%
- Government consumption and investment
- 16.5%
Source: US Bureau of Economic Analysis
Three estimates, one quarter
The BEA publishes the same quarter three times. The advance estimate arrives about a month after the quarter closes and is built partly on assumptions where source data is missing. The second and third estimates replace those assumptions with actual data as it arrives. Revisions between them are routinely large enough to change the narrative.
Beyond that come annual updates each summer and comprehensive revisions every five years, which can rewrite the level and the growth path of years already in the history books. GDP is not a measurement taken once; it is an estimate that converges.
Because of the lag, markets lean on nowcasts - models that assemble the monthly data as it arrives into a running estimate of the current quarter. They are useful and volatile, and they are not the BEA's number.
- Advance estimate: about a month after the quarter ends.
- Second and third estimates: at roughly monthly intervals after that.
- Annual updates each summer; comprehensive revisions every five years.
- Nowcasts are models, not official statistics, and they move on every monthly release.
Behind the subscription
The rest of this entry is the part that changes a decision: what moves the price, which contract sets it, who ships it and where that can be cut off.
What moves a print
The seven components behind a GDP surprise - and the two of them that regularly produce a headline that says the opposite of what the economy did.
Where a growth view is priced
The instruments that carry a growth view, the nowcast that trades as a preview, and why no exchange lists GDP itself.
From transactions to a recession call
The path from monthly data to a dated recession, and why the committee that dates them takes a year to do it.
How this shows up in prediction markets
Which vintage settles a GDP question, why recession markets hang on a committee rather than a number, and the checks before you price either.
Included with a subscription
Create an account to unlock the full entry — price drivers, trading venues, trade flows and the live markets attached to it.
Frequently asked questions
- Is two negative quarters of GDP a recession?
- Not officially in the United States. Recessions are dated by an academic committee that weighs employment, income, production and sales, and it has declared recessions without two negative quarters and declined to declare them after two. The rule of thumb is journalism, not a definition.
- Why is US GDP growth annualised?
- Convention. The US reports what the quarterly change would compound to over a year, so a 0.5% quarter is published as roughly 2%. Most other countries publish the plain quarterly change, which makes unconverted comparisons meaningless.
- Why does GDP get revised so heavily?
- The first estimate is published a month after the quarter ends, before much of the source data exists, so parts of it are assumption. Later estimates replace assumptions with data, and annual and five-yearly revisions can rewrite entire years.
- Why can a strong economy produce a weak GDP print?
- Because imports are subtracted and inventories are counted as production. A quarter of strong demand met by imports, or a quarter in which firms sell down stock they built earlier, can produce a weak headline while underlying demand is fine.
Primary sources
Related entries
Prediction markets carry real risk of loss. Nothing on Market Guy is financial advice — it is research tooling to help you think, not a signal to trade.