The One Manager Skill That Persists
Hierarchical orthogonal decomposition applied to manager selection — of style timing, sector timing, and stock picking, only stock picking persists out of sample.
Research
Methodology, empirical studies, and the One Position, Four Bets series — hierarchical orthogonal decomposition applied to positions, 13F books, and mutual-fund panels.
The One Position, Four Bets series and empirical working papers first; shorter product and methodology notes below.
Series parts and working papers — figures, samples, and replication detail.
Hierarchical orthogonal decomposition applied to manager selection — of style timing, sector timing, and stock picking, only stock picking persists out of sample.
Turning conviction into tradeable risk: same label, different bets across AAPL/NVDA, XOM/KMI, and MAG7.
Market, thematic, and stock-specific risk across Buffett, Ackman, Lone Pine, Tiger Global, and Baupost — and what survives the 45-day filing lag.
When-issued market mechanics, the CRSP convention, and how RiskModels incorporates spin-off pricing — the IBM/Kyndryl case
A within-mandate manager-efficiency framework using ERM3 residual decomposition
Subsector ETF value, joint optimization, and executable hedge layers across 9,074 US mutual funds
Top cohort holdings refresh, factor decomposition, and residual winners for the March 2026 reporting window
Early ownership, active conviction, and residual attribution in U.S. mutual-fund managers, 2019–2026
Top-decile rank persistence, active-share comparison, and tail-stratified inference across 1,000 top-AUM US mutual funds
Shorter explainers on API surfaces, screening, and how to read the cascade.
Vasicek peer-β cross-sections expose what sector ETFs paper over
How RiskModels picks the right hedge depth automatically, per stock, per day
Server-side rank screening turns the universe into one queryable cross-section
Part 1 — The problem. Custodial reporting obscures concentration. Four names that looked diversified drew down 50%+ together in 2022. Style factors (Growth, Value) are symptoms, not drivers — subsectors are the real unit of risk.
Part 2 — The manager. The same decomposition applied to five concentrated 13F filers: how portfolio risk partitions across market, thematic, and stock-specific layers; how active structure compounds in dollars; and what survives a realistic filing lag.
Part 3 — The one skill that persists. The same hierarchical orthogonal decomposition (Market → Sector → Subsector → style), now applied to manager selection and tested out of sample across diversified US equity funds: of style timing, sector timing, and stock picking, only stock picking carries forward — a durable, between-fund skill an allocator can rank on.
Run the decomposition logic in the browser — no account required.
RiskModels ecosystem
RiskModels.org stays the credibility layer: methodology, proof, and exhibits. Product links are kept contextual so the research remains the primary object.
Research
Methodology, article series, and public exhibits for institutional review.
Read the researchAPI
REST API, SDKs, CLI, and MCP-ready endpoints for reproducible decomposition calls.
Open API docsWorkspace
Web application surface for portfolio workflows, dashboards, and authenticated product use.
Open web app