After a first cash payout, do next-month returns differ between recurring and one-off payouts?
Study 29 · Inconclusive · recorded 1 October 2026 · answer: Not settled · full paper with figures
Correction, 2 October 2026. The text of this record was rewritten to the project's editorial standard, from the study's plan, result and records rather than its earlier wording. What the study found is unchanged.
The question
A first payout here follows a year without one, and whether it recurs is known only after 400 more days, so nobody could act on it on the day. We compare returns net of the market over the 21 sessions after the payment date, for payouts that recurred and payouts that did not.
Rules fixed before the result
Before any return was computed, we fixed the two groups, the 21-session window and the test: the difference in median returns counts only if it falls outside the middle 95% of 20,000 random relabellings of the same payouts. At least 40 one-off payouts from 30 companies and 80 recurring payouts were required. A difference would have made a follow-up on payout type worth planning; none means no follow-up on this record.
Data and method
The outcome is the sum of a share's daily returns minus a broad market index's over the 21 sessions after the payment date. Recurring means three or more further payouts on a schedule within 400 days; one-off means no further payout in that time, and irregular payers are left out. The measure is the recurring group's median minus the one-off group's median.
The answer
Inconclusive: across 200 payouts, the median return of recurring payers was 2.3 percentage points above that of one-off payers, inside the no-effect range of minus 3.1 to plus 3.3 points that 95% of random relabellings of the same payouts produce, as we expected. Two of ten alternative measures, the difference of means and the return without market adjustment, each fell above its own no-effect range, both favouring recurring payers. Both were fixed in advance as checks that do not decide.
Main estimate
0.023 fraction (95% range -0.031 to 0.033), from 200 observations
What it changes
Nothing changed. No difference was found at this sample size, so no follow-up study on payout type is planned on this record.
What remains uncertain
The groups are known only after 400 days, so nobody could have used them on the day. The window starts at the payment date, not the announcement, and before the study we expected most of any difference to fall before that date, where this study does not look. Keeping only each company's earliest payout removed the difference entirely.
How it was made
Run took 19 minutes; independent code review: approve; the recorded calculation reruns to the same result.
Fingerprints
plan ffbb7d302be05940a3a94b53e182e8061656d2cf75cbe0a11b8cbda6a2e6691a result 1af12d1238366417faeb3de509b0885ed2266e578c5257c55bdf602c43bbcafd
Exploratory work on development data; not a performance claim.