Defined inputs
Reject inputs published after a decision cut-off.
Research programme 05 / time-aware finance
What was knowable at the time?
Financial as-of evidence, timestamp provenance and replayable decision boundaries. Synthetic examples expose look-ahead bias; they do not establish market returns.
0 public research assetsEvidence-bound demonstrations
INVARIANT / OFIRPublic experience
As-of evidence · No look-aheadSynthetic visualization
03 / THE WORKBENCH
Portfolio Explorer
A new synthetic interactive teaching/prototype exercise. It is not part of the verified 130-package archive.
Starting the original interactive module…
Synthetic examples only. No market return claim or production trading system.
01 / RESEARCH QUESTION
Does the tested strategy use information not yet published?
Compare observation, publication and evaluation dates. Flag rows unavailable at the chosen decision time.
Each result is constrained to a stated experiment. Programme hypotheses and independent validation are not interchangeable.
02 / METHOD & BOUNDARY
What can be checked now
A research site, not a claim of final proof.
Explicit comparison
As-of Evidence Workbench · synthetic timestamp audit
Do not infer
Synthetic examples only. No market return claim or production trading system.
OPEN SCIENCE / STATUS
OFIR is building its dedicated evidence track.
There is no separately listed OFIR repository among the 108 currently published research assets. This site demonstrates a synthetic time-aware finance method, not verified forecasting performance.