Does NQ come back to its 7:00 open?
NQ, one-minute bars, 2008-01-02 to 2026-08-14. The reference is the open of the 7:00 bar; the reading is the 8:59 close. A day counts when the reading is at least 0.10% away from the reference. Returned means some bar from 9:00 through 3:59 touched or traded through the reference. Beside it, the same test on the 8:00, 8:30 and 9:30 opens read two hours later, and on a price picked at random from the 6:00 to 8:59 morning, read at 9:00 like the 7:00 open.
What the file says
Any morning price a tenth of a percent away gets revisited about three days in four. NQ came back to its 7:00 open on 2,274 of 3,073 days, 74.0%. A price drawn at random from the morning, read at the same 9:00 and given the same session to return in, was revisited on 1,968, 76.0%. Started below the open and returned up, 75.8%; started above and returned down, 72.3%. On a two-year window a shared version of this stat was cut on, the 7:00 open returns 70.3% and the random morning price 73.9%: the chosen open is not the best of the references, it is one of them.
Five references, the same test
The white rows are the pair to compare: both read at 9:00 with the whole session to return in.
| reference | days | returned | started below, returned | started above, returned |
|---|---|---|---|---|
| the 7:00 open | 3,073 | 2,274 · 74.0% | 1,136 of 1,498 · 75.8% | 1,138 of 1,575 · 72.3% |
| the 8:00 open | 3,813 | 2,216 · 58.1% | 1,086 of 1,833 · 59.2% | 1,130 of 1,980 · 57.1% |
| the 8:30 open | 3,984 | 1,971 · 49.5% | 993 of 1,934 · 51.3% | 978 of 2,050 · 47.7% |
| the 9:30 open | 4,076 | 1,574 · 38.6% | 761 of 1,898 · 40.1% | 813 of 2,178 · 37.3% |
| a random price from the 6:00 to 8:59 morning | 2,588 | 1,968 · 76.0% | 1,000 of 1,293 · 77.3% | 968 of 1,295 · 74.7% |
The same, on the last two years only
2024-09-09 to 2026-08-14: the window a shared version of this stat was cut on.
| reference | days | returned | started below, returned | started above, returned |
|---|---|---|---|---|
| the 7:00 open | 340 | 239 · 70.3% | 116 of 158 · 73.4% | 123 of 182 · 67.6% |
| the 8:00 open | 401 | 213 · 53.1% | 106 of 192 · 55.2% | 107 of 209 · 51.2% |
| the 8:30 open | 429 | 187 · 43.6% | 99 of 208 · 47.6% | 88 of 221 · 39.8% |
| the 9:30 open | 424 | 134 · 31.6% | 62 of 192 · 32.3% | 72 of 232 · 31.0% |
| a random price from the 6:00 to 8:59 morning | 280 | 207 · 73.9% | 102 of 129 · 79.1% | 105 of 151 · 69.5% |
By year, the 7:00 open
| year | qualifying days | returned | share |
|---|---|---|---|
| 2008 | 206 | 152 | 73.8% |
| 2009 | 187 | 146 | 78.1% |
| 2010 | 169 | 132 | 78.1% |
| 2011 | 173 | 134 | 77.5% |
| 2012 | 165 | 122 | 73.9% |
| 2013 | 143 | 94 | 65.7% |
| 2014 | 123 | 89 | 72.4% |
| 2015 | 149 | 112 | 75.2% |
| 2016 | 135 | 95 | 70.4% |
| 2017 | 107 | 74 | 69.2% |
| 2018 | 170 | 131 | 77.1% |
| 2019 | 155 | 118 | 76.1% |
| 2020 | 183 | 140 | 76.5% |
| 2021 | 165 | 120 | 72.7% |
| 2022 | 209 | 165 | 78.9% |
| 2023 | 179 | 130 | 72.6% |
| 2024 | 169 | 115 | 68.0% |
| 2025 | 168 | 124 | 73.8% |
| 2026 | 118 | 81 | 68.6% |
By year, the 7:00 open ran from 65.7% (2013) to 78.9% (2022); every year of nineteen is above 65%.
What it does not say, and how it was done
What this does not measure. The path: touched again by 3:59 says nothing about how far price went first, where a stop would have sat, or what the trade cost. And one line of the table measures the clock, not the reference: the 8:00, 8:30 and 9:30 opens are read two hours later and so have less of the session left to return in (five hours, four and a half, four and a half against the 7:00 open's seven); their lower shares are the shorter day, and they are kept here because they were declared. The fair comparison for the 7:00 open is the random morning price, read at the same 9:00.
How it was measured. The measurement, its references and six predictions were written down before it ran (SETUP 03, September 10, 2026; four of the six held, one miss being the clock effect above, and one the stat coming in stronger on eighteen years than the band allowed). Script nq_open_return-2026-09-10.py v0.1 on the one-minute NQ file: 6,180,656 lines, 4,811 dates with morning bars, 2008-01-02 to 2026-08-14. The random draw is one 6:00 to 8:59 close a day, seed 3. The script's own print is the record for this page, nq_open_return.txt; every number here is read from it.
Raw market data: FirstRate Data (firstratedata.com). A measurement on a file, not advice.