An automated equity research pipeline. It screens a broad universe of US-listed
companies daily, scores every name against a fixed model, and publishes a ranked
shortlist with the evidence behind it. It places no trades and gives no personalised
advice.
The pipeline
Evidence flows one way. The risk review can exclude a candidate; it can
never promote one or reorder the ranking.
What subscribers receive
Every published pick is tracked after the fact and reported back —
including the ones that did not work. The weekly digest is open signup; the conviction
list and index levels are currently sent on request.
Operating principles
One list for everyone
Every subscriber gets identical research at the same time. Nothing is tailored to
an individual.
Fixed model
Scoring does not adapt itself. Revisions are evidence-based, reviewed, and
released deliberately.
Judgement can only subtract
Rankings are mechanical. Review may exclude a name, never elevate one.
Measured in public
Published picks are scored against benchmarks and reported — wins and losses
alike.
Fails safe
If an optional stage is unavailable, the pipeline continues without it rather than
publishing something unverified.
Research only
Fully separated from any trading system. It holds no positions and places no
orders.
Get the research free
The weekly sector digest arrives by email — sign up in a few seconds. The Top-5 conviction list and daily index
levels are not yet public: reply to any issue to be added.
Scope. A high-level overview of how the research is produced.
Data providers, signal definitions, model weights, ranking and screening rules are
proprietary and are not described here.