A research process designed to resist self-deception
Ideas are defined before observation, tested prospectively, reconciled independently, and advanced only through explicit review.
Follow a sleeve from raw data to approval
Five layers stand between a hypothesis and broker execution. Every one is designed to expose weak evidence early.
Collect point-in-time market data
Prices, SEC filings, classified news, macro series and insider activity are captured with the date each fact became public -- so a strategy can only ever be judged on what was knowable at the time.
Freeze the strategy before judging it
A strategy is written down as fixed rules -- what it looks for, how it enters, how it protects, how it exits -- before any test begins. Frozen means the test cannot quietly become the tuning.
Shadow every decision without touching money
New ideas run on paper first: real data, real decisions, zero capital. Analyst opinions are frozen nightly and cannot place, size or veto anything.
Grade against independent truth
Results are measured against the broker's records and each name's own benchmark -- never against the strategy's opinion of itself. Self-graded edges have fooled this firm before; the rule exists because of it.
The Gate: evidence in, most ideas out
Only a measured record with honest sample sizes can move an idea forward -- and the owner signs every capital decision. The kill record is kept on purpose: what failed, and why, stays published internally.
The first conversation is about process, not promises.
Reviews begin with source quality, risk, and the complete evidence record—including decisions that did not work.