After a one-day fall that a company filing dates, does a third or more of the further decline happen overnight?
Study 74 · No reading (stopped by its own checks) · recorded 29 October 2026 · answer: Not settled · full paper with figures
The question
After a sharp one-day fall that a company filing dates, shares of the smaller listed companies the project follows kept falling over the next three sessions. We ask whether a third or more of that further decline happens in the two overnight gaps rather than in trading hours, each share compared with quiet shares of similar price and volatility.
Rules fixed before the result
Fixed before any outcome: the answer reads the 95% range of the two overnight gaps' summed return, net of comparison shares, against zero and 0.50 percentage points, about a third of an earlier study's decline. A range holding zero and wholly below 0.50 meant under a third overnight; wholly above zero but starting below 0.50, a visible part; starting at 0.50 or more, a third or more. The study first needed 6,000 kept falls, agreeing price constructions and enough precision.
Data and method
As planned, each fall's two overnight gaps over the three sessions after it were to be compared with quiet shares' gaps on the same days, in the same price group (a share price above or below ten) and volatility tenth and summed. Prices the source adjusts for splits and payouts were to decide, with an independent adjustment of its unadjusted prices as the check. The 95% ranges were to allow for falls on neighbouring dates and repeat companies.
The answer
Not measured: the study stopped at its first step, before any return after a fall was formed. A check fixed in advance required the unadjusted closing price, divided by the split-adjusted closing price, to stay constant between recorded share splits on 1,000 sampled shares; on some it jumped as a split would with no split recorded, so the independent second construction of the overnight gaps could not be relied on, and by its own rule the study recorded no answer. The kept falls also numbered fewer than the 6,000 the plan required, because the rule that sets aside any fall near an event of an earlier study removed the falls priced at ten or more. The question is answered neither yes nor no.
What it changes
No plan changed: by the rule set in advance, a stop at this step leaves the record with its counts, and a corrected design is for a later choice. The account under test, that selling by holders in trading hours carries the further decline, is neither supported nor contradicted here.
What remains uncertain
Nothing after a fall was measured, so neither answer is shown. These falls were studied before, so even a measured answer would have been a check within our development data, not a test on fresh data. Daily prices cannot separate selling at the opening auction from news absorbed overnight, and trading volume was only ever a side condition, never the test.