Correction, 30 September 2026. Technical wording rewritten in shorter sentences under the readability rule; the meaning is unchanged.
Correction, 1 October 2026. The outcome label now follows the study's value record, which counts this study as having changed a rule or a plan.
Correction, 19 October 2026. This text was reviewed after the study finished, in a catch-up of the earliest studies; the result, outcome and dates are unchanged.
Rules fixed before the result
The rules for each possible answer were written down and fixed before any result was seen.
What remains uncertain
Whether each company's ordinary returns cluster by date, so that sessions near an event differ from its whole history, is still open at one and three months. Only one pool of split events was used, the one earlier methods studies used, and at 63 sessions overlap between the same company's windows was not separated from that clustering. Whether a real split session resembles its company's nearby or whole-history sessions cannot be seen in a methods study.