Among the least traded shares, do a week's winners still trail its losers when screened on the traded price?
Study 70 · No reading (stopped by its own checks) · recorded 26 October 2026 · answer: Not settled · full paper with figures
The question
An earlier study found that among the third of shares that trade the least by value, priced at five or more in the market's own currency, a week's winners trailed its losers the next week. Its screen read prices adjusted for later payouts and share changes, letting in shares that traded below five; this check uses the day's traded price.
Rules fixed before the result
Fixed before any outcome: the earlier study's rules unchanged except that its price screen of five or more reads the price as traded. The answer is whether the gap under that screen has its whole range below zero; the weekly difference of the two screens' gaps measures how much the adjusted price carried, and decides only when that range holds zero. A week counted if both screens kept at least 200 shares, and 200 such weeks were required.
Data and method
As in the earlier study, shares in the third that trades the least by value are sorted at each week's last session by the week's return, the top fifth against the bottom fifth from the next week's first opening price, each week's outcomes capped at each week's own 1st and 99th percentiles, not removed. Before any return was formed, the check counted how many shares each screen kept week by week; the years 2013 to 2017 set the expected precision.
The answer
Not measured: the check stopped at its first step, before forming any return in the years tested. Its rules, fixed before any outcome, required at least 200 weeks in which both the earlier screen and the traded-price screen kept at least 200 of the least traded shares; the traded-price screen leaves out the shares that traded below five, and too few weeks kept enough, so by its own rule the study stopped there. The question is answered neither yes nor no: whether the earlier gap holds on the traded price, and how much of it the shares let in only by the adjusted price carried, are neither shown nor ruled out. The plan had given a clear answer a chance of 0.86 and a stop of this kind, or a failed rebuild, a chance of 0.05.
What it changes
No plan changed and nothing is corrected: the earlier finding stands as published, with its stated limitation that its price screen read adjusted prices still open. Whether a corrected design for this check is written is decided at the next review.
What remains uncertain
Nothing in the years tested was measured, so what the traded-price screen does to the earlier gap is unknown. The years 2018 to 2022 were used by the earlier study and others, so even a measured answer would have been a check within our development data, not a test on fresh data. The adjusted prices are the vendor's as of its snapshot, and only one choice was varied.