Opening gaps: how often the prior close is touched, against the mirror level
- Gap filled (prior close touched)
- Mirror level touched (same distance, other side)
Hover or tap a pair of columns.
October 17, 2026 · 3 min read
Do Gaps Fill? NQ and ES Opening Gaps Against a Fair Base Rate
NQ fills its opening gap on 61% of sessions and ES on 59%. A level the same distance on the other side of the open gets touched almost as often. 7,400 sessions of gap statistics, compared with the base rate.
"Gaps fill" is one of the most repeated rules in futures trading. It is usually stated as a percentage, with no comparison. We measured it on every regular session since 2013 for NQ and since 2011 for ES, then checked it against the test it needs: how often does price reach a level the same distance from the open in the other direction?
The headline number
Across 3,476 NQ sessions that opened away from the prior close, price traded back to that close before the 4:00 pm close on 61.3% of them. ES filled on 59.4% of 3,940. So gaps do fill more often than not. The question is whether that has anything to do with the gap.
The base rate: the mirror level
Take a session that gaps up 0.4%. The prior close sits 0.4% below the open. The mirror level sits 0.4% above it. If gaps pull price back toward the prior close, the fill should be touched clearly more often than the mirror. If the fill is just a level near the open, both should be touched at about the same rate.
| Gap size | NQ sessions | NQ fill | NQ mirror | ES sessions | ES fill | ES mirror |
|---|---|---|---|---|---|---|
| Under 0.25% | 1,361 | 83.8% | 83.5% | 1,812 | 81.3% | 77.4% |
| 0.25–0.5% | 873 | 60.4% | 58.4% | 1,030 | 53.7% | 53.1% |
| 0.5–1% | 791 | 44.4% | 38.1% | 742 | 32.6% | 35.8% |
| Over 1% | 451 | 24.8% | 20.8% | 356 | 20.5% | 14.9% |
| All gaps | 3,476 | 61.3% | 58.7% | 3,940 | 59.4% | 57.6% |
The mirror is touched 58.7% of the time on NQ and 57.6% on ES. The gap adds about two to three points over a level chosen only for its distance from the open. Most of the famous fill rate is the size of a normal day's range.
Is the extra two or three points real?
Possibly, but only just. Counting the sessions where only one of the two levels was touched, the fill wins 890 to 801 on NQ (z = 2.2) and 990 to 916 on ES (z = 1.7). By size bucket the picture is uneven. NQ's 0.5–1% bucket shows the largest gap, +6.3 points. ES's same bucket goes the other way, −3.2. With eight buckets across two markets, one or two cells reaching z ≈ 2 is about what chance produces.
What the gap adds: fill rate minus mirror rate, in percentage points
Hover or tap a bar. Right of zero, the gap fills more often than the mirror level is touched.
What does change the fill rate: size and time
- Size dominates. Gaps under 0.25% fill 81–84% of the time on the two markets, and their mirrors are touched 77–84% of the time. Above 1%, fewer than one in four fill.
- Small gaps fill early. The median fill time is 9:33 am ET on NQ and 9:37 on ES for gaps under 0.25%, against 11:57 and 11:52 for gaps over 1%.
- Down gaps fill a little more often than up gaps (NQ 63.8% vs 59.4%), but their mirrors are also touched more often (60.3% vs 57.6%). Down days have wider ranges.
- The rate drifts between eras. NQ filled 64.1% of gaps in 2013–2018 and 59.2% in 2019–2026, and the mirror fell by the same amount.
Why a fill rate is not a trade
A fade that targets the prior close wins when price touches it. It loses on the sessions that run the other way, and those can run a long way, with nothing in the touch rate to say how far. With the mirror level touched almost as often as the fill, a fade with a symmetric stop is close to a coin toss before costs. Whether a particular version works depends on its stop, its entry time and its size filter, and those are questions for a backtest, not a percentage.
Data and method. E-mini Nasdaq-100 (NQ) and E-mini S&P 500 (ES) futures, 1-minute regular-session bars, NQ 2 January 2013 and ES 3 January 2011, both to 8 October 2026. The gap is the 9:30 am ET open against the previous regular-session close, in the same contract, on consecutive sessions; roll days are left out. A gap fills when price trades back to the previous close by the 4:00 pm close. The mirror level is the same distance from the open, on the other side. The paired comparison is a McNemar test on sessions where exactly one of the two levels was touched.