GDP and PCE on 20 February 2026: separate the release, unit and vintage

A dated correction distinguishes quarterly SAAR GDP from monthly PCE, records later GDP revisions and withdraws unsupported expectations and influencer outcomes.

What changed in this article

Editorial correction · 9 September 2026. The original article was published on 21 February 2026 and combined two separate Bureau of Economic Analysis (BEA) releases from 20 February. It treated a quarterly annualized GDP rate, a monthly PCE price change and an unverified expectation comparison as one scorecard. The retained page also did not include the call register, benchmark series or calculations behind its 64-call hit rate and return claims. Those comparisons and outcomes are withdrawn. This correction preserves the original URL and publication date and reports what each release actually measured.

The two original-vintage releases

Release	Reference period	Original-vintage fact	Measurement boundary
GDP (Advance Estimate), 4th Quarter and Year 2025	2025 Q4; released 20 February 2026	Real GDP increased 1.4% at an annual rate; the Q4 PCE price index increased 2.9% at an annual rate and the PCE price index excluding food and energy increased 2.7%	A quarterly change reported at a seasonally adjusted annual rate (SAAR), not a December monthly rate
Personal Income and Outlays, December 2025	December 2025; released 20 February 2026	The PCE price index increased 0.4% from November and 2.9% from December 2024; excluding food and energy, it increased 0.4% month over month and 3.0% year over year	A monthly price-index change; current-dollar PCE also increased 0.4%, but spending and prices are different series

Both reports say their releases were rescheduled because of the October–November 2025 government shutdown. That is part of the release chronology, not evidence that the estimates were free of later revisions. The GDP release says the advance estimate was originally scheduled for 29 January; the PIO release gives the same original date for December data.

The original page’s 0.4% monthly and 3.0% core year-over-year values are therefore recognizable at the stated precision, but its “0.3% expected,” “about 2.7% expected,” “about 2.9% expected” and surprise labels are not retained: no contemporaneous expectations snapshot was supplied. The original page’s 2.8% headline year-over-year figure also does not match the December 2025 PIO release, which reports 2.9%.

Later vintages changed the GDP reading

BEA’s second estimate, released 13 March 2026, revised Q4 real GDP down to 0.7% SAAR, a 0.7 percentage-point revision from the advance estimate. Its Q4 PCE price index and core measure remained 2.9% and 2.7% SAAR. The third estimate, released 9 April 2026, revised Q4 real GDP again to 0.5% SAAR; the two Q4 PCE measures remained 2.9% and 2.7%. These later releases are not substituted for what a reader could have known on 20 February; they are labeled later vintages.

The January 2026 PIO release, released 13 March, still displays December’s PCE price-index changes as 0.4% month over month and 0.4% excluding food and energy at the shown precision, while noting that estimates for earlier months were updated. That is a reminder to record the vintage and release date when a historical comparison is rerun.

What is withdrawn

The retained page did not provide a source-complete register for the claimed 64 calls, 14 of 64 correct, 21.9% hit rate, −1.7 percentage-point three-day result, cluster returns, hedged-baseline comparison or the claimed influencer-portfolio causal path. Those results are not verified findings and are withdrawn rather than recalculated from current data. The employment and policy-rate entries in the original table are also not treated as part of the two BEA releases.

This correction does not establish that a soft landing was confirmed or broken, that a particular portfolio underperformed, or that the release caused a specific yield or equity move. It establishes only the dated observations and their frequency, annualization and vintage boundaries.

A reproducible next question

To test whether a public macro call anticipated a release, a future register would need the release identifier and timestamp, series and reference period, SAAR/monthly/year-over-year unit, frozen expectation source, call timestamp and wording, instrument and venue, execution convention, benchmark, fees and the later-vintage rule. Without those fields, a narrative can be discussed as a hypothesis but not scored as a hit rate or return study.

Original publication: 21 February 2026. Correction: 9 September 2026. AI-assisted source checking and writing; no independent human expert review. This page is educational market research, not personalized investment advice.