Do speculative shares fare worse against safe shares from open to close after a weekend than before it?
Study 63 · Inconclusive · recorded 21 October 2026 · answer: Not settled · full paper with figures
The question
Every trading day from 2018 to 2022 we ranked shares priced at five or more on four traits known the day before, and took the most speculative fifth's return from the open to the close less the safest fifth's. We compared the first session after each weekend with the last session before it, one difference per weekend.
Data and method
Each trading day, shares priced at five or more were ranked on four traits known at the previous close: volatility over the last sixty-three sessions, a low price, the largest daily gain of the last twenty-one sessions and low trading value. The spread was the most speculative fifth's return from the open to the close less the safest fifth's. Each weekend gives one difference, the first session after it less the last session before it.
The answer
Not settled: over 260 weekends the gap averaged 0.16 percentage points lower on the first session than on the last before it (95% range -0.32 to 0.01), so that size is neither shown nor ruled out; the median weekend gave -0.10 (-0.28 to -0.003). Overnight, from close to open, it went the other way, 0.12 (0.04 to 0.19). Taking out each fifth's sensitivity to the market's daily move times the market's weekday gap left -0.15 (-0.29 to -0.01). It was -0.27 (-0.42 to -0.12) among the third of shares that trade the least by value and 0.11 (-0.13 to 0.36) among the most traded third; a Tuesday less Thursday placebo gave 0.16 (-0.04 to 0.37). The speculative fifth trailed the safest by 0.21 on first sessions (-0.34 to -0.08) and by 0.05 on last sessions (-0.17 to 0.07).
Main estimate
-0.159 percentage points (95% range -0.324 to 0.007), from 260 observations
What it changes
No plan changes: by the rule set in advance, an unsettled result records its range, adds no weekday control to our other studies of speculative shares and proposes no follow-up study. The plan had expected a gap of about half to two thirds of the 0.24-point gap seen on 2013 to 2017 before the plan was fixed, with low confidence, and had given a result like this one a chance of 0.27.
What remains uncertain
The weekday pattern is already documented, and study fifty-seven read the years 2018 to 2022 for other questions, so this is a check within our development data, not a fresh test. The market sensitivities came from 2013 to 2017, a thinly traded share's opening price can come minutes into the session, and the volatile years 2020 to 2022 widened the range. We hold no measure of investor mood, so a mood effect is judged only from where the gap appears.
How it was made
Run took 95 minutes; independent code review: approve; the recorded calculation reruns to the same result.
Fingerprints
plan b8ca723434ee119f0373ab99e8612ab3e0e67880d02f3c70f2db3e08e4d0356b result 52db23e6fdca97838f5d0ef95859cbc77a3d29426675b6fe050461d9c7fede21
Exploratory work on development data; not a performance claim.