Gap-Open Regime Study
Forward intraday paths after significant gap opens, by group & regime · October 29, 2026 · 4:03 PM ET

What happens after a significant gap — hour by hour, over the next 10 sessions

The backdrop: oil and a growth-slowdown note gapped the index lower, semis & memory much more, and a bid ran into energy & refiners. This study pools every gap open (open vs prior close) that clears a group's threshold, then chains the cumulative forward return in 60-minute windows across the next 10 sessions. Each group reads inside its own regime — semi / AI / index since Oct 1 2025, energy since Mar 1 2026. The landing line is absolute; toggle to relative to QQQ (name minus the index backdrop). The heatmap is the median cumulative return at each session × time-of-day cell. Tested direction: down-gaps (index −1.25% · tech ETFs −3% · tech names −5%) test the bounce; up-gaps (energy +3%) test the continuation. Mirror lens: the same universes and regimes with the thresholds flipped — tech / index gapping up (+1.25 / +3 / +5%) tests continuation, energy gapping down (−3%) tests the bounce — so a morning like Sep 16 (semis up into strength, energy down after the run) gets a forward-path read too.

Live triggers · 2026-10-29 — 2 in the tested direction, 0 in the mirror · click any name for its chart & trade brief: MRVL -11.57% brief › PBF +4.74% brief ›
▼ Gap-down · semi / AI / index · since Oct 1 2025

Index backdrop

since Oct 01, 2025 gap ≤ -1.25%
39 gap events across 4/4 names · forward path over 10 sessions
bounced (10-sess)
59%
Median 10-sess
+1.0%
Per-name setups · best probability first
▸ show the 39 individual events

Semi / AI ETFs

since Oct 01, 2025 gap ≤ -3.0%
67 gap events across 4/4 names · forward path over 10 sessions
bounced (10-sess)
57%
Median 10-sess
+2.25%
Median vs QQQ
+1.28%
Per-name setups · best probability first
▸ show the 67 individual events

Semi / AI single names

since Oct 01, 2025 gap ≤ -5.0%
128 gap events across 15/16 names · forward path over 10 sessions
bounced (10-sess)
50%
Median 10-sess
-0.02%
Median vs QQQ
+0.34%
Per-name setups · best probability first
▸ show the 128 individual events
▲ Gap-up · energy · since Mar 1 2026

Energy ETFs

since Mar 01, 2026 gap ≥ +3.0%
1 gap events across 1/5 names · forward path over 10 sessionsthin n<5
kept rising (10-sess)
0%
Median 10-sess
-7.41%
Median vs QQQ
-18.99%
Per-name setups · best probability first
▸ show the 1 individual events

Energy single names

since Mar 01, 2026 gap ≥ +3.0%
33 gap events across 10/13 names · forward path over 10 sessions
kept rising (10-sess)
64%
Median 10-sess
+2.81%
Median vs QQQ
-0.37%
Per-name setups · best probability first
▸ show the 33 individual events

Energy futures (outright)

since Mar 01, 2026 gap ≥ +3.0%
20 gap events across 4/4 names · forward path over 10 sessions
kept rising (10-sess)
70%
Median 10-sess
+5.05%
Per-name setups · best probability first
▸ show the 20 individual events
Alpha Query · Gap forward-return study

—

Method. Gap = open ÷ prior close − 1. For each qualifying event, cumulative return is measured from the event-day open at each regular-hours 60-minute bar (09:30–15:30 ET, ~7 per session) over the next 10 sessions; futures are restricted to the same RTH window so they align with the index. Paths are pooled by window index; the band is p25–p75. Relative = the name's path minus QQQ's over the same clock. Energy futures are outright (CL/RB/HO/BZ) — calendar spreads are a follow-up. Small samples (n<5) are flagged; read them as directional, not precise. Educational research, not advice.