Release study // Bureau of Economic Analysis
Personal income and outlays (PCE): how far markets move after the release
Personal income and outlays (PCE): across 126 releases since 2016, the Nasdaq-100 fund (QQQ) typically moved 10.8 bp in either direction in the 30 minutes after the release, against 9.3 bp in the control window (n = 24).
The typical size of the move
Over these 126 releases, against 24 control mornings, the typical 30-minute move in the Nasdaq-100 fund was 1.2 times the size of the control window's. In either direction: the study measures how much prices moved, not which way.
By fund and window
Basis points of each fund's price. Typical size is the median move with its sign removed; the middle half is the 25th to 75th percentile of the signed moves. Each window counts its own releases: a morning is left out for a fund that did not trade before a measuring point. The control is the same thirty minutes on the business day before (there is no five-minute control) and leaves out days that were themselves a release morning at that time.
| Fund and window | Typical size after | Releases (n) | Typical size, control | Control mornings (n) | Middle half after | Median, signed |
|---|---|---|---|---|---|---|
| Nasdaq-100 fund QQQ, 5 minutes | 5.5 bp | 126 | Not measured | · | -4.1 bp to +8.4 bp | +0.3 bp |
| Nasdaq-100 fund QQQ, 30 minutes | 10.8 bp | 126 | 9.3 bp | 24 | -9.2 bp to +11.7 bp | +1.0 bp |
| S&P 500 fund SPY, 5 minutes | 4.2 bp | 126 | Not measured | · | -3.2 bp to +6.0 bp | +0.6 bp |
| S&P 500 fund SPY, 30 minutes | 8.1 bp | 126 | 3.0 bp | 24 | -8.4 bp to +8.0 bp | -0.1 bp |
| Long Treasury fund TLT, 5 minutes | 9.8 bp | 110 | Not measured | · | -7.8 bp to +11.2 bp | 0.0 bp |
| Long Treasury fund TLT, 30 minutes | 19.1 bp | 110 | 5.8 bp | 22 | -16.3 bp to +21.0 bp | +2.9 bp |
The last 12 releases
Each row is one release morning (n = 1 per row): the signed move in basis points from just before the release. Listed by date, not sorted or colored by result.
| Release date | QQQ, 5 minutes | QQQ, 30 minutes | SPY, 5 minutes | SPY, 30 minutes | TLT, 5 minutes | TLT, 30 minutes |
|---|---|---|---|---|---|---|
| -26.0 bp | -32.7 bp | -16.4 bp | -20.4 bp | -16.8 bp | -28.7 bp | |
| -0.6 bp | +22.8 bp | -6.3 bp | -4.3 bp | -9.7 bp | -19.3 bp | |
| +25.2 bp | +24.3 bp | +20.9 bp | +18.3 bp | +17.3 bp | +33.3 bp | |
| +9.6 bp | +32.7 bp | +6.7 bp | +19.1 bp | +12.9 bp | +32.9 bp | |
| -12.0 bp | -3.9 bp | -10.9 bp | -6.3 bp | -10.5 bp | +9.3 bp | |
| -3.1 bp | -3.6 bp | -2.4 bp | -0.3 bp | -11.5 bp | -13.8 bp | |
| +0.7 bp | +8.5 bp | +1.1 bp | +7.9 bp | +12.7 bp | +30.0 bp | |
| -32.8 bp | -36.3 bp | -22.9 bp | -21.5 bp | -4.5 bp | -16.7 bp | |
| -3.2 bp | +13.0 bp | -0.4 bp | +10.6 bp | -5.7 bp | 0.0 bp | |
| -0.7 bp | -19.4 bp | -0.4 bp | -15.2 bp | -24.0 bp | -34.3 bp | |
| -9.5 bp | -28.2 bp | -6.1 bp | -18.5 bp | -6.9 bp | -15.8 bp | |
| +22.6 bp | +18.2 bp | +15.5 bp | +15.9 bp | -6.7 bp | -1.1 bp |
How this is measured
We measured every personal income and outlays release from February 1, 2016 to August 26, 2026, 127 release mornings. On each one we take three exchange-traded funds, QQQ (the Nasdaq-100), SPY (the S&P 500) and TLT (long-dated Treasuries), from the last one-minute price before the 8:30 a.m. ET release to five and to thirty minutes after it. The control is the same thirty minutes on the business day before each release, leaving out a day that was itself a release morning at that time in our history. Figures are in basis points (hundredths of a percentage point): the median and the middle half (25th to 75th percentile) of the moves with their sign, and the median size of the move whichever way it went. Each window shows how many mornings it covers. We hold no consensus forecast, so no release is sorted into a beat or a miss: these figures describe how large the reaction usually is, not which way it goes. They describe what happened before and do not predict the next release.
- Each figure is a size, not a direction. We hold no consensus forecast, so we do not sort releases into beats and misses, and a large move after one release says nothing about the direction of the next one.
- Moves are in basis points of the fund's price: 100 basis points is 1%. The typical size is the median of the moves with their sign removed, so half of the measured releases moved less and half moved more. The middle half runs from the 25th to the 75th percentile of the signed moves.
- The control window is the same thirty clock minutes on the business day before each release, for the same funds, leaving out a day that was itself a release morning at that time. It is a baseline for that hour of the day, not a morning with nothing scheduled: the notes say what it still holds, such as releases we do not measure yet. Every figure carries the number of mornings it rests on (n).
- What happened before is not a forecast. Information only, not financial advice.
Notes
- A morning is left out for a fund when that fund did not trade in the ten minutes before a measuring point, which happens in thin pre-market trading, so the counts differ by fund and window.
- On 22 of these 127 mornings GDP came out at the same time, so those mornings measure both releases together.
- The control is the same clock window on the business day before, not the minutes just before the release: a window just before a 10:00 a.m. release would include the 9:30 a.m. open, the busiest half hour of the day.
- On 101 of these 127 mornings the business day before was itself a release morning at 8:30 a.m. (GDP on 101), so the control leaves those days out and rests on the other 26.
- We have no stored history yet for CPI, PPI and retail sales, which also come out at 8:30 a.m., so a control day that fell on one of them is still counted.
- The weekly jobless claims usually come out at 8:30 a.m. ET on Thursdays, and we do not measure them: 26 of these 127 release mornings and 6 of the 26 control mornings fell on a Thursday, so those windows hold that release as well.