Hypothesis graveyard
What we tested — and what failed. The normal output of our research is a null; the permanent record of rejected hypotheses is the honest result.
Pre-registered studies: the hypothesis and its success criterion are committed before the result is computed. "Rejected" means the hypothesis failed its own pre-committed test — that is a result, not a failure of the method.
Pre-registered studies
Every study was registered in writing before the run; the document is auditable in the repository.
| Study | Registered | Status | Result |
|---|
Factor Discovery Program
Candidates are judged per family with a Benjamini-Hochberg FDR; rejected hypotheses are kept permanently. Promotion is rare and requires an explicit sign-off. Nothing here is a signal, a direction, or a recommendation.
| Family | Tested | Rejected | Survivors |
|---|
The graveyard
Permanent. Rejected hypotheses are never resurrected.
| Hypothesis | Family | Rejected | p (out-of-sample) | FDR-corrected |
|---|
· Research classification only. Not investment advice, no recommendation, no performance claim.