Momentum Regime-dependent component

API endpoint: GET /companies/{ticker}/momentum

A ROC20/ROC60-based momentum score, ported from the source trading bot. Tested cross-sectionally: at each monthly rebalance date, sort the tracked universe into deciles by momentum score and measure each decile's forward return, at 4 horizons (20/60/120/252 trading days).

The real result

Mean rank correlation is negative at every horizon tested (20d: -0.055, 60d: -0.049, 120d: -0.024, 252d: -0.015) - shrinking toward zero with distance, not flipping positive. This is a real, if weak, mean-reversion signal, not the classic academic momentum-continuation effect.

Quality scorecard

Dimension	Score
Sign consistency	69.8%
Signal strength	0.038
Horizon consistency	100% (all 4 horizons agree on sign)
Crisis robustness	87.5%
OOS stability	27.4%

Why "regime-dependent," not "core factor": sign consistency and OOS stability both fall short of our core-factor bar. Crisis robustness rising to 87.5% is consistent with a well-documented academic effect ("momentum crashes" - high-momentum stocks tend to suffer sharply after a stress event), not a contradiction.

Query it

curl -H "X-API-Key: <your-key>" \ https://finance.ordinexdata.com/companies/AAPL/momentum

Full validation methodology: FACTOR_VALIDATION_METHODOLOGY.md.

Free early access to the API: Get API access →