After a two percent rise, do shares closing near the day's high beat those closing low from the next open?
Study 64 · Changed a rule or plan · recorded 21 October 2026 · answer: Yes · full paper with figures
The question
On sessions from 2018 to 2022 when a share priced at five or more the day before rose two percent or more, we compared those that closed in the top fifth of the day's range with those that closed in its lower half. Both groups were measured from the next session's open to its close.
Rules fixed before the result
Fixed in advance: at least 500 sessions with both groups, an average of 1,000 shares a session, and a size of 0.16 percentage points, a gap being present only if its range lies above zero and reaches it. The weighted average decided unless its range came out wider than set; then the median over sessions, if trials on 2013 to 2017 with signs shuffled by week and run in streaks showed few false alarms. It was precise enough and decided.
Data and method
The two groups were compared inside each session, which removes the market's common move that day but not the groups' different sensitivity to it, and sessions were averaged with each weighted by how many shares it had in the two groups. The range allows for dependence across nearby sessions. After falls of two percent or more the mirror gap was -0.04 (-0.15 to 0.07), and sorting all shares by where they closed gave -0.05 (-0.15 to 0.04) over five sessions.
The answer
Yes: over 1,251 sessions, shares closing near the high earned 0.22 percentage points more from the next open to the close than those closing low (95% range 0.15 to 0.30), reaching the 0.16-point size fixed in advance. Close to close the gap was -0.13 (-0.23 to -0.04), as overnight it was -0.36 (-0.43 to -0.30), consistent with a close at the asking price given back at the open. It held among the most and least traded halves by value, 0.11 (0.001 to 0.22) and 0.28 (0.19 to 0.37). Allowing for the rise's size, the day's range, volatility, market sensitivity and trading value left 0.18 (0.10 to 0.26). On their own, from the next open to its close, near-high closers averaged -0.03 (-0.11 to 0.05), like every eligible share's -0.04 (-0.10 to 0.03), and low closers -0.30 (-0.41 to -0.19).
Main estimate
0.222 percentage points (95% range 0.147 to 0.297), from 1251 observations
What it changes
By the rule set in advance, as the gap held after allowing for the rise's size, the day's range, volatility, market sensitivity and trading value, our records propose that studies of returns from the next open after a big rise allow for where a share closed. The gap was 0.13 (0.04 to 0.21) points larger among risers than among all shares. The plan expected a gap in both trading halves (chance 0.27), and both showed one.
What remains uncertain
The closing bounce and overnight reversal are known effects, and four earlier studies used 2018 to 2022, so this is a check on already-used data, not a fresh test. Daily prices cannot tell a close given back at the open from overnight news, nor when the gap is earned. A thin share's opening price can come minutes late, a price floor checked the day before admits shares below it that day, and early 2020's turmoil may dominate the result.
How it was made
Run took 110 minutes; independent code review: approve; the recorded calculation reruns to the same result.
Fingerprints
plan 27e2c553124af65ba072920d77b609cb45e78100b5f9a94b9bf07df347d7997e result 0f76f0aae5ca3965b384d5d7d159fed133c3ca707e5c5e1fc15608b639cd4cb6
Exploratory work on development data; not a performance claim.