Guide · Markets
Annualised vs Year-Over-Year Growth: Reading GDP
An annualised rate shows what a quarter's change would add up to over a full year, while a year-over-year rate compares with the same period a year earlier. Here is why the US could report 3.4% growth and 3.6% inflation on the same morning.
Economic data are reported as growth rates, and the same underlying numbers can produce very different headlines depending on which rate is used. The two most common are the annualised rate and the year-over-year rate. Mixing them up is one of the easiest ways to misread a release, and the US reports of October 29, 2026 put both side by side.
An annualised rate takes the change from one quarter (or month) to the next and asks what it would amount to if it continued for a full year. US GDP is reported this way. If output rises 0.8% from one quarter to the next, the annualised rate is roughly (1.008 × 1.008 × 1.008 × 1.008) − 1, about 3.2%; the Bureau of Economic Analysis works from unrounded data, so a quarter reported as 0.8% can be published as 3.4% annualised, as happened in the third quarter of 2026. The advantage is speed: it shows the economy's momentum in the latest quarter. The drawback is noise, because any one-off swing in a single quarter is multiplied by four.
A year-over-year rate compares a period with the same period a year earlier: September 2026 against September 2025, or the third quarter against the third quarter of 2025. It is smoother and is how most inflation figures, including the Fed's 2% target, are expressed. But it is slow, because it includes changes that happened up to twelve months ago, so it can keep rising after momentum has already turned, or fall while prices are speeding up again.
The October 29 releases show the difference. Real GDP grew at an annualised 3.4% in the third quarter of 2026, up from 2.2%, which came from a quarterly increase of about 0.8%. In the same report, the PCE price index rose at an annualised 2.1% in the quarter, down from 5.0%, which looked like inflation was nearly back to target. Yet the monthly report released at the same time showed PCE prices 3.6% higher than a year earlier in September, and core prices 3.1% higher. Both were correct: the second quarter's energy shock was in the year-over-year number but not in the third quarter's quarter-on-quarter change, while September alone rose 0.5%, a pace that would be over 6% if it lasted a year.
Monthly data can be annualised the same way, by compounding over twelve months. A 0.3% monthly rise in core prices is about 3.7% annualised, and a 0.2% rise about 2.4%, which is why economists say "two-tenths a month is roughly consistent with the Fed's target". Some series, such as housing starts, new home sales and car sales, are published as seasonally adjusted annual rates (SAAR): the monthly figure is multiplied up to show what a full year at that pace would look like, not what was actually sold in the month.
When you read a release, check three things before reacting: which rate the headline uses, what period it compares, and what the other rate says. A strong annualised quarter after a weak one can leave year-over-year growth unchanged; a soft quarterly inflation print can coexist with a rising annual rate. Markets react to surprises against forecasts, and forecasts are quoted in the same convention as the release, so the right comparison is always like with like.