GDP measures final production within an economy over a period. A real-output estimate, a nominal value and an annualised growth rate answer different questions. This article explains those distinctions, the US estimate sequence and the revisions to inspect when comparing a release with expectations.

What GDP actually measures

GDP, gross domestic product, measures final domestic production without double counting intermediate goods and services. Nominal GDP reflects both prices and quantities; real GDP adjusts for price changes. Rising nominal GDP alone therefore does not establish increased real output.

BEA publishes quarterly and annual US GDP estimates. Use the release’s own period, units and revision status. Countries have different publication schedules and conventions; a US quarterly headline should not be treated as a global definition of GDP.

GDP rises without one additional unit

Build a one-final-good economy. It sells 50 units in each of two periods. Only the price rises in the second period.

Scroll sideways to see the full diagram

Nominal growth without quantity growthThe good's price rises from 2 to 2.20. Quantity is 50 units in both periods. Nominal value rises from 100 to 110; value at the base price stays 100.SAME QUANTITY, HIGHER PRICE50 units in both periods; price 2 → 2.20Nominal valueAt base price 2100A110B100A100BNominal: 100 → 110, change +10%Real in this model: 100 → 100, change 0%

Hypothetical economy: one good produced and sold domestically, no intermediate goods, imports, taxes or other changes. The constant base price is 2 units. This is a simple model, not a reproduction of BEA's chain-weighted method. BEA: nominal and real GDP supports the distinction between current-price value and quantity after adjusting for price changes.

Nominal bars differ in height; constant-price bars are identical.

Before comparing headlines, check whether they describe current-price value or real production growth. In our model the 10% value increase comes entirely from price. It does not establish a currency reaction.

Three estimates of the same quarter

US quarterly GDP normally has advance, second and third estimates as additional source information becomes available. These are successive vintages, not guarantees of accuracy, and the published schedule can change. The third estimate can still be revised in later annual or comprehensive updates.

  1. The advance estimate — the first scheduled quarterly estimate, using information available at that time. Some inputs are estimated because source data are incomplete. Check the actual release date rather than assuming a fixed lag or market reaction.
  2. The second estimate — updates the same quarter using additional information. The revision can be upward or downward; its market impact is not predetermined by its position in the publication sequence.
  3. The third estimate — incorporates further source data. It is not based on every possible final input and does not close the revision process: subsequent updates can change the estimate.

Read the date and time on BEA’s current release calendar and convert Eastern Time for that date into your own zone. Publication times and daylight-saving offsets should not be treated as year-round constants. The euro area and the United Kingdom follow their own schedules; the US estimate sequence is not a global rule.

Why it is quoted at an annual pace

US GDP is reported in a form that trips up many beginners: as a seasonally adjusted, annualised quarter-on-quarter rate. It sounds complicated, but the idea is simple. The bureau takes the change over a single quarter and scales it to an annual pace — it shows how fast the economy would grow over a full year if it kept the momentum of those three months.

In a hypothetical real-output index, a rise from 100 to 100.5 is 0.5% quarter on quarter. Annualising gives ((100.5 / 100)4 − 1) × 100 = about 2.0151%. This compounds the same quarterly pace four times; it is neither a forecast nor year-on-year growth. Year-on-year growth requires the level one year earlier. Read each release’s units: the annual-rate convention is not exclusive to the United States.

Consensus and GDPNow are different estimates

GDP covers a period already past, and some relevant monthly indicators are available before its release. It is nevertheless not a simple sum of jobs, sales and survey headlines. BEA combines many sources, with adjustments and estimates for missing information, to measure final domestic production.

Atlanta Fed GDPNow updates a model estimate as source data arrive, without discretionary adjustments. It is not an official Atlanta Fed or FOMC forecast. A survey consensus combines respondents’ forecasts according to the provider’s method; save its timestamp and check whether it is a mean or median. See how to read the economic calendar for recording expectations before publication.

Hypothetical comparisons, all in real quarter-on-quarter growth at an annual rate
ComparisonValuesDifference
Release versus consensus2.5% versus 3.0%−0.5 percentage points
Revised versus earlier previous quarter1.8% versus 1.5%+0.3 percentage points
Release versus a model estimate2.5% versus 2.2%+0.3 percentage points

These are three distinct comparisons. Do not add them together or interpret any of them as a percentage change in a currency pair.

Interpreting growth and its components

A surprise or revision may change expectations, but it does not prescribe a dollar move. Inspect consumption, investment, government spending, net exports and inventories; distinguish a component’s growth rate from its contribution to GDP growth in percentage points. Compare policy expectations for both currencies rather than inferring a trade from the headline alone.

Imports illustrate why components need context. In a hypothetical fixed-price expenditure model, an imported final consumer good worth 100 adds 100 to consumption and 100 to imports: the net change in C + I + G + X − M is zero. If the buyer pays 110 and the import costs 100, with 10 of domestic distribution value added, the net contribution is 10. Removing the first purchase reduces both C and M by 100; it does not automatically raise GDP by 100. This is level accounting under the stated assumptions, not a calculation of chain-weighted contributions to real-GDP growth.

A GDP result can change a policy view, but it does not determine the Fed’s decision or a currency direction. Two successive negative real-GDP quarters are a common shorthand, not the NBER’s US recession-dating rule, which considers a broader decline in activity. Read GDP alongside the Fed rate decision and ISM manufacturing index, within the wider fundamental analysis framework.

What to do at the next GDP release

  1. Note the date and time in your own zone. Check the BEA release calendar for the next estimate and record its vintage. Convert the published Eastern Time for that date into your local zone; do not use a year-round fixed time difference.
  2. Record the consensus and nowcast before the release. Save their timestamp, period and units. Agreement between two estimates does not establish a small probability of surprise; note the uncertainty and compare the later result without hindsight.
  3. Separate the annual pace from the quarterly change. Copy the exact units from the release. Check whether you are comparing annualised quarter-on-quarter, unannualised quarter-on-quarter or year-on-year growth before calculating a surprise.
  4. Log the release and reaction. Record expectations, the announced figure, revisions and the observation window for the dollar. A few quarters are observations to review, not proof that GDP usually confirms rather than reverses trends.