207 / 300 fully matched non-quarantined decisions. Minimum gate not reached.
EXPERIMENT H2 / H2_SPOOL_PIT_V3R2
Does Korea lead Binance?
We prospectively test whether information from Korean exchanges adds incremental lead-lag value at the moment a Binance candidate is generated.
STATUSEFFECT BLINDEDManual review status: BLOCKED
BLOCKED. Current blocker: minimum_fully_matched_nonquarantined.
PASS / STALE. Experiment version, quarantine revision, clock checks and preregistration hash are consistent in the latest gate snapshot.
4 decisions are excluded from the confirmatory cohort while their original records remain preserved.
BLINDED. No automatic effect opening. Passing maturity only permits manual review.
LIVE OFF. No automatic execution change is permitted from this research arm.
Candidate→PIT Capture→Endpoint Contract→Quarantine→Maturity→Manual Review
WHY THIS MATTERSEven if it's the same second
Even if it's the same second
It could be the future.
Cross-venue research must distinguish event time, receive time, first-seen availability, and decision time. V1 was frozen for confirmatory use because it did not satisfy that contract rigorously enough. V3R2 therefore uses only samples collected after a new prospective start.
Verified from the research state at 2026-10-03 00:20 UTC. Public pages intentionally omit exact scoring formulas, actionable trading thresholds, live candidates, API credentials and sensitive infrastructure details. Upstream exact raw lineage remains partially UNKNOWN for some earlier V3R2 decisions.
COMMUNITY NOTES
Join the discussion →What are the strongest counterarguments to this hypothesis?
Share alternative interpretations of the data conditions, timing assumptions, selection bias, or normalization choices.