# Clustered insider purchases: two negative event-information trials

**Short title:** Clustered insider purchases: complete trial lineage  
**Author:** Arhan Canli, Founder and Quantitative Researcher, Canli Capital  
**Family key:** `equity_insider_activity` · **System:** ALPHAC / AlphaForge  
**Status:** public research record; not peer reviewed; not an investment solicitation

## Finding and boundary

The mechanism asks whether purchases by multiple officers or directors reveal shared private
belief about undervaluation. Two charged implementations produced annualized Sharpe -0.9931 and
-0.2433 over 2,669 observations. Both are negative. The literature prior includes Lakonishok and
Lee's study of insider trading information
([Review of Financial Studies](https://doi.org/10.1093/rfs/14.1.79)); it does not validate ALPHAC's
Form 4 timestamps, clustering, universe, costs, or borrow assumptions.

The identities belong to one event-information family and cannot become two sleeves. A related
probe artifact is hash-bound, but identity-matched curves, drawdowns, DSR, capacity, and continuous
broker evidence remain incomplete. **Decision: FAIL / zero sleeves.** Exact evidence is in
[`equity_insider_family.json`](/glassbox/equity_insider_family.json). The preregistered work and
system were authored and directed by **Arhan Canli**; no Alpaca return is attributed to it.