Do shares follow their closest-moving peers' past week after the weekend, but not between ordinary weekdays?
Study 59 · Inconclusive · recorded 19 October 2026 · answer: Not settled · full paper with figures
The question
Prices of related shares moving one after another, a lead-lag, are a documented kind of effect. Our scans of the years 2013-2017 saw shares whose closest-moving peers fell most over a week trail those whose peers rose most on the first trading day after the weekend, and not between ordinary weekdays; we tested it on the years 2018-2022.
Data and method
A share's closest-moving peers are the fifteen whose daily returns, net of the market-wide move, tracked it most closely over the prior 250 trading days, re-chosen every twenty-one trading days, using past returns alone. Each week, among shares priced at 5.00 or more in US currency, we compared the top and bottom fifths by the peers' week return on the next trading day's return against the market, over 208 weeks, beside ordinary weekdays and fifteen placebo shares of middling co-movement.
The answer
Not settled. In 2018-2022, on the first trading day after the weekend, the top fifth by the closest-moving peers' past week beat the bottom fifth by 0.19 percentage points on the clipped average, the measure we fall back on (95% range -0.06 to 0.43), and 0.22 on the plain average (-0.03 to 0.48); both ranges hold zero and the 0.33 of 2013-2017. Allowing for weeks sharing peers, it was -0.07 to 0.45. Ordinary weekdays gave 0.01 (-0.08 to 0.10), placebo shares -0.01 (-0.05 to 0.04), and heavily traded shares with heavily traded peers 0.10 (-0.19 to 0.39). The peers moved together; overall the next day followed mostly their last trading day, the pattern the plan reads as stale closing prices, yet the slope on their whole week, with each share's own returns held, stayed above zero (0.022; 0.0015 to 0.043).
Main estimate
0.002 fraction (95% range -0.001 to 0.004), from 208 observations
What it changes
By the rule set in advance, our measures of returns after events followed by a weekend add no peer-group control, and the next peer-group question waits until we have looked at why the range stayed wide. The plan expected a catch-up after the weekend only, with low confidence, at 0.20 to 0.40 points.
What remains uncertain
These years' prices were read before by studies eighteen, fifty-seven and fifty-eight for other questions, so this is a check within our development data, not a fresh test; the year 2018 served only to choose the first peers. The peers are chosen by co-movement alone, with no industry data; weeks that share one month's peers are not fully independent; shares are equally weighted; and the price filter uses the week's last close. The holiday cells are too thin to read.
How it was made
Run took 121 minutes; independent code review: approve; the recorded calculation reruns to the same result.
Fingerprints
plan bd11ca91c26e5f8f09a7f2283a278859027de340086b647af3f65bf8dacce480 result 079c43e1612a6be5ea4f348d1f0c5d5634bc9f067d5715094fc9fba91048c1cd
Exploratory work on development data; not a performance claim.