Anomaly Discovery Pipeline
Purpose
Research on manager universes often stops at the paper. In a data-driven sales architecture, the same dataset should continuously surface commercial opportunities: pain points, peer underperformance, and allocator-relevant anomalies.
How it works
- Frozen eVestment manager-month panels across six universes (quintiles, persistence, skill evolution, regime cuts)
- YAML-driven auto-research loop that ranks marketable anomalies
- Output: candidate topics for insight notes, sales alerts, and targeted consultant outreach
Example use case
Fundamental small-cap pain points surfaced as ranked candidates, connecting quantitative manager-universe research to a specific distribution action rather than a static research archive.
Strategic value
Shared data foundation for three major papers and the sales targeting layer. Research and distribution read from the same observable opportunity set.