ReturnSkew

Return Skewness — Bali Engle and Murray 2015

Risk

Quick facts

Factor ID ReturnSkew
Display name Return Skewness
Family Predictor
Category Risk
Direction lower_expected_return
Status supported
Paper Bali Engle and Murray 2015
Source tables daily
Required fields ret
PIT required no
Coverage 2010-02-03 – 2026-05-15 · 679,322 rows · 99.2% of panel

Formula

skew(ret over t-21..t-1)

Sample skewness of daily returns over the prior trading month using raw-moment identities.

China adaptation

Computed on raw not market-adjusted returns to match the original Bali-Engle-Murray construction.

Direction-adjusted cumulative return

1.00 1.56 2.12 2.69 3.25 2010-02 2014-02 2018-03 2022-04 2026-04

Return metrics

Metric Value
Months in sample 195
Monthly mean 0.5423%
Monthly std 2.1809%
Annualized Sharpe 0.86
t-stat (simple) 3.47
Hit rate (months > 0) 62.6%
Last 12M compounded -13.36%
Cumulative wealth (start = 1) 2.742

Recent 12 months

Month L/S return Cumulative Names
2025-05 -1.55% 3.116 5129
2025-06 -1.25% 3.077 5140
2025-07 -1.56% 3.029 5143
2025-08 -1.75% 2.976 5147
2025-09 +1.27% 3.013 5150
2025-10 -1.17% 2.978 5152
2025-11 -3.09% 2.886 5162
2025-12 +0.52% 2.901 5167
2026-01 +0.17% 2.906 5178
2026-02 +0.55% 2.922 5180
2026-03 -3.36% 2.824 5187
2026-04 -2.89% 2.742 5174

Known data issues

Daily price limits clip the upper tail of returns and may compress skewness for limit-hitting names.

Notes

Closely related to lottery-demand factors and overlaps mechanically with MaxRet.

See also: Risk family, Factor library, Factor returns.