SUMP5 SUMP

Sum of positive changes over total change

Formula

Sum(Greater($close-Ref($close, 1), 0), 5)/(Sum(Abs($close-Ref($close, 1)), 5)+1e-12)

The expression above is verbatim from qlib Alpha158/360 feature engineering (qlib/contrib/data/loader.py, Apache-2.0). The platform qlib-compatible expression engine evaluates it on the deterministic research panel.

Cross-section statistics & information coefficient

IC mean (60M)

-0.0099

ICIR

-0.179

Cross-section mean

0.5200

Cross-section std

0.3149

Min

0.0000

Max

1.0000

IC = Spearman-style rank correlation between the factor cross-section and next-month returns, computed over the last 60 months on the deterministic research panel.

Data export

alpha158.json

Related factors · SUMP

SUMP10

Sum(Greater($close-Ref($close, 1), 0), 10)/(Sum(Abs($close-Ref($close, 1)), 10)+1e-12)

SUMP20

Sum(Greater($close-Ref($close, 1), 0), 20)/(Sum(Abs($close-Ref($close, 1)), 20)+1e-12)

SUMP30

Sum(Greater($close-Ref($close, 1), 0), 30)/(Sum(Abs($close-Ref($close, 1)), 30)+1e-12)

SUMP60

Sum(Greater($close-Ref($close, 1), 0), 60)/(Sum(Abs($close-Ref($close, 1)), 60)+1e-12)

Alpha158 — Alpha158 (Microsoft qlib) — 158 features… Factor Library Factor Hub Model Lab