Dataset
ProFactor Library
Cross-sectional factor value coordinates for Taiwan equities (five families), each with definition and formula version — coordinates only, no signal.
What this data is
Each row is one security, one trading day, one factor, carrying factor_value (the raw coordinate), factor_family (size / value / momentum / quality / low_vol), factor_definition, and formula_version, with as_of_source_date marking the point-in-time input date. Factor values are COORDINATES only — no signal, score, or buy/sell label. Keyed on trade_date and point-in-time safe.
Use cases
- Build cross-sectional stock selection and factor models on consistently defined coordinates.
- Align formula_version and as_of_source_date for reproducible, look-ahead-free backtests.
- Pair with Factor Returns to assess each factor's historical decile spread.
Why it matters for stock analysis
Home-grown factors drift in definition, convention, and point-in-time alignment; taking coordinates that carry their definition, formula version, and input date keeps research consistent and look-ahead-free.
Coverage / Freshness / Source Policy
Per ticker, per trading day, per factor, across the size / value / momentum / quality / low-vol families; every row carries factor_definition, formula_version, and as_of_source_date. Keyed on trade_date, point-in-time safe.
Daily, updated with the cross-sectional inputs.
Derived from official TWSE / TPEx / MOPS sources; factor values are coordinates, not signals, not investment advice.
Developer entry points
- API docs: /docs/api/structure-reference/factor-library
- OpenAPI: /openapi.json