Before the cutoff, after it, after publication: three kinds of seasonal result
Every trade after a research cutoff is kept out of the figures. Not every one of them tests anything. The dates decide which do.
A rule's record can carry three kinds of result, and they deserve different weight. The history runs to the research cutoff, 31 December 2025. After it, SPY's turn-of-the-month and FOMC-day rules completed 10 and 6 trades, but every one of them closed before 17 October 2026, the day these rules were published as research topics. So none is a forecast tested in the open. The first that will be is the FOMC-day trade entered on 28 October 2026.
History, after the cutoff, and after publication
The three are easy to blur, because all of them are trades of the same rule, priced the same way.
- The history. Every year up to the research cutoff, 31 December 2025. These are the years the rule's figures were computed on, and the years anyone choosing among rules could see. Nothing in them tests a choice.
- After the cutoff. Trades whose prices came later than the cutoff. The evaluator keeps them apart from every research figure. But a trade that had already happened when someone chose the rule was visible to that person, so it is not a clean test either.
- After the rule was fixed in public. Trades that start after the rule, with its exact definition, was published. Only these are results nobody could have seen when the rule was chosen.
A research cutoff marks where the data ends. It does not record when a rule was picked, so a cutoff alone is not a registration.
The dates that decide which claim a trade supports
Every completed trade so far sits before the rules were published.
scan_20261004T193904Z, and the repository's record of when the topics were published.| Date | Event | Source |
|---|---|---|
| 2025-12-31 | Research cutoff: the last day any research figure can use | Evaluator field research_cutoff |
| 2026-10-04 | The evaluator run behind these figures finished | scan_20261004T193904Z, finished_at |
| 2026-10-16 | Latest price in the snapshots | Evaluator field data_as_of |
| 2026-10-17 | The two rules published as public research topics | Research-pages release, project log |
| 2026-10-19 | This guide's figures retrieved and frozen | Snapshot provenance |
The run date matters too. The evaluator run finished on 4 October 2026; 15 of the 16 trades had closed by then, and 1 closed between the run and publication. A rule picked after looking at that run could have been picked with those trades in view.
Every completed trade after the cutoff, by claim
Both rules' trades after the cutoff are listed with the claim each can support. As of 16 October 2026, the column never reads “after the rule was public”.
| Entry open | Exit open | Excess | Which claim |
|---|---|---|---|
| 2025-12-30 | 2026-01-06 | −0.08 pp | Priced before the evaluator run |
| 2026-01-29 | 2026-02-04 | −1.02 pp | Priced before the evaluator run |
| 2026-02-26 | 2026-03-04 | −1.83 pp | Priced before the evaluator run |
| 2026-03-30 | 2026-04-06 | +2.31 pp | Priced before the evaluator run |
| 2026-04-29 | 2026-05-05 | +1.36 pp | Priced before the evaluator run |
| 2026-05-28 | 2026-06-03 | +0.90 pp | Priced before the evaluator run |
| 2026-06-29 | 2026-07-06 | +1.51 pp | Priced before the evaluator run |
| 2026-07-30 | 2026-08-05 | +5.26 pp | Priced before the evaluator run |
| 2026-08-28 | 2026-09-03 | −0.65 pp | Priced before the evaluator run |
| 2026-09-29 | 2026-10-05 | +0.22 pp | Priced after the run, before the rule was public |
| Entry open | Exit open | Excess | Which claim |
|---|---|---|---|
| 2026-01-28 | 2026-01-29 | −0.13 pp | Priced before the evaluator run |
| 2026-03-18 | 2026-03-19 | −1.74 pp | Priced before the evaluator run |
| 2026-04-29 | 2026-04-30 | +0.47 pp | Priced before the evaluator run |
| 2026-06-17 | 2026-06-18 | −0.25 pp | Priced before the evaluator run |
| 2026-07-29 | 2026-07-30 | −0.57 pp | Priced before the evaluator run |
| 2026-09-16 | 2026-09-17 | +0.44 pp | Priced before the evaluator run |
The turn-of-the-month trades after the cutoff averaged +0.80 pp of excess (a plain mean of the 10 listed, derived), far above its history. That is worth noticing and not worth much: ten trades, already priced before anyone fixed the rule in public. The turn-of-the-month research keeps them apart from its history for that reason.
Where the clean record starts
The FOMC-day rule's next entry is 28 October 2026; the turn of the month's is 29 October 2026. Both fall after publication, so their results, whatever they are, will be the first of each rule that nobody could have seen when the rule was fixed. One trade will say little; a record built this way is slow by design.
The research pages list these trades as they close, beside the history and never inside it.
What the dates do not tell you
Publication fixes a rule's definition; it does not make the rule good, and a clean record of a few trades is still a few trades. The dates here are the ones this repository can show. If a rule was chosen earlier than its publication, the clean record could start earlier, but without a dated record of that choice these pages do not claim it.
This guide's figures were frozen on 19 October 2026. Trades that close after that date are on the research pages, not here.
The track record follows the registered strategies after the research cutoff in the same way.
Watch the first trades after publication
Each research page adds its trades after the cutoff as they close, kept apart from the history.
Open the FOMC-day researchSources and research notes
- Evidence: SeasonalEdge public research API,
/api/v1/research/turn-of-the-month-spyand/api/v1/research/fomc-announcement-day-spy, retrieved 2026-10-19T20:51:46Z; rule ids92688fa1e33982f6,35e591dac79d6da6; evaluator runscan_20261004T193904Z, finished 2026-10-04T21:22:06. - Publication date: the public research topics were released on 17 October 2026 (the research-pages release, recorded in the project's task log). Earlier choice dates are not claimed.
- Derived: each trade's claim is set by comparing its exit date with the run and publication dates; the mean of the turn-of-the-month trades after the cutoff is a plain mean of the listed trades.
- How SeasonalEdge research is computed · Glossary
Join our newsletter
We share our research, product news and upcoming seasonalities. Free, and you can unsubscribe at any time.