anomaly-atlas
Honesty doctrine. Every candidate anomaly is an artifact until proven otherwise; in-sample results are never findings; past statistical regularity does not imply future returns. This is research on statistical properties of market data — not investment advice, not a trading system.

Code / experiments/micro/expD_leadlag_scan/README.md

experiments/micro/expD_leadlag_scan/README.md 16 lines
---
project: anomaly-atlas
document: expD_leadlag_scan/README
author: Simon-Pierre Boucher
contact: contact@spboucher.ai
data_source: hfmarketdata.io
created: 2026-08-12
status: draft
---

# expD_leadlag_scan

Lead-lag scan: lagged cross-correlations and Granger, separating timestamp artifacts

Status: **completed 2026-08-12** — decay 2006-07→2014-15 measured on synchronized pairs; ES↔SPY splice-invariant basis effect; SPX lead = content staleness (H11 confirmed at scan level).