Does NQ break both ends of the premarket hour more on Mondays?

The 72% is real. Monday is not the reason. Both ends of the 8:12 to 9:12 hour were traded through by the close on 616 of 870 Mondays over eighteen years, 70.8%, and on 71.2% of the last twelve months’ Mondays. Tuesday is 70.8% as well, and Wednesday 69.7%; the week falls to Friday’s 59.2%. What moves with the number is the width of that hour: a median 13 points on a Monday and 20.88 on a Friday, inside a day range near 64 points either way. A narrow box inside a normal day gets broken both ways more often. Monday beat the other four days in 13 of 19 years and lost 6.

Both ends broken counts a break from 9:12, the minute after the range closes; from 9:30 only counts breaks after the bell. Neither end broken: 9 days of 4,559. Which end broke first is close to even on every weekday: the high 51.0%, the low 48.7%.

Points, median over the file. The day’s range barely moves across the week; the hour’s width does, and the double-break rate moves the other way with it. Friday’s hour is the widest, a third of the day, and Friday is the day both ends break least.

The file’s last twelve months, 2025-08-15 to 2026-08-14. Mondays 71.2% against the shared 72%: the stat reproduces on recent data. Over eighteen years Tuesday matches it.

White rows are the years Monday beat the other four days: 13 of 19. Mondays ran from 50.0% (2019) to 82.7% (2018). 2026 runs to 2026-08-14.

Of the 3,056 double breaks on the file, the first end broke at a median 9:28, before the bell: 1,702 of them (55.7%) had their first end broken between 9:12 and 9:29. The second end broke at a median 9:51; by 10:30 on 73.0%, by 11:00 on 80.5%, by noon on 88.0%, by 2:00 on 94.7%.

What this does not say. It is a count of days, not a trade. Trading it, break one end and aim for the other, needs a stop, a target, costs and a clock, and is its own test. The 8:30 releases sit inside the hour on many Fridays, one reason that hour runs wide; the measurement does not separate them. Roll days are in. MNQ is the same prices.

How it was done. The question, the definitions and six predictions were written down before anything ran (SETUP 12, September 29, 2026; six of six held). nq_tbr_double_break-2026-09-29.py v0.1 read the one-minute NQ file: 6,180,656 lines, 0 unparsed, 4,811 dates with bars from 8:12 to 4:59, 2008-01-02 to 2026-08-14. 4,559 days measured; 10 set aside for no 9:30 to 3:59 session and 242 for a missing minute inside the 8:12 to 9:11 hour, most of them in the thin early years. The range is the high and low of the 60 bars stamped 8:12 through 9:11; a break is a bar’s high above the range high or its low below the range low, by the 3:59 bar. The print, nq_tbr_double_break.txt, and one row per day, nq_tbr_double_break.csv, are the record for this page; every number here is read from the print.

Raw market data: FirstRate Data (firstratedata.com). A measurement on a file, not advice.