High convictionsigma:04 · n=27
Stretched factor — mean-reversion watch: Quality - Extended Risk d1
Quality - Extended Risk d1 = +2.55% (+3.83σ vs 60d)
3.83σ extended vs 60d (n=27). Factor and regime extremes are risk-STATE signals, not direction calls (ref: the -3.45σ Momentum-LowVol week).
ExpressExpress via the pair: fade the leg that ran — long Extended Momentum Risk / short Quality / Low-Vol Uptrend (or trim the spread).
FramingA >=2σ dislocation has historically reverted more often than it extends — this is a fade-the-extreme probability, not a directional forecast.
InvalidationInvalidated if the print keeps widening on rising volume or a fresh regime break confirms the move.
High convictionregime:04 · n=194
Regime context — size to the state: Constructive Rotation
Constructive Rotation (composite 0.66, confidence 0.80)
This regime historically persists ~5 td, p_stay 0.80. Labels are monotone in forward VOLATILITY, not direction — Distribution Risk historically preceded the BEST fwd-20d returns (+2.54%).
ExpressSet gross/net to the volatility state, not the label's tone; keep position sizing modest here.
FramingA risk-state gauge: size positions, do not chase direction off the label.
InvalidationInvalidated on a confirmed regime transition (the two-run confirmation rule).
Medium convictionsigma:04 · n=27
Stretched factor — mean-reversion watch: Extended Momentum Risk d1
Extended Momentum Risk d1 = -2.12% (-2.16σ vs 60d)
2.16σ compressed vs 60d (n=27). Factor and regime extremes are risk-STATE signals, not direction calls (ref: the -3.45σ Momentum-LowVol week).
ExpressTrim or rotate exposure in the named basket; do not add into the extreme.
FramingA >=2σ dislocation has historically reverted more often than it extends — this is a fade-the-extreme probability, not a directional forecast.
InvalidationInvalidated if the print keeps widening on rising volume or a fresh regime break confirms the move.
Medium convictionsigma:04 · n=60
Stretched factor — mean-reversion watch: Credit w1
Credit w1 = +1.01% (+2.08σ vs 60d)
2.08σ extended vs 60d (n=60). Factor and regime extremes are risk-STATE signals, not direction calls (ref: the -3.45σ Momentum-LowVol week).
ExpressExpress via the pair: fade the leg that ran — long LQD / short HYG (or trim the spread).
FramingA >=2σ dislocation has historically reverted more often than it extends — this is a fade-the-extreme probability, not a directional forecast.
InvalidationInvalidated if the print keeps widening on rising volume or a fresh regime break confirms the move.
Low convictiongap:01 · n=74435
Gap fade — down-gap bounce into the close
GVA gapped -13.95% at the open
Down-gaps >=1% have bounced +13-14bps from 10:00 into the close (n=74,435, t=14.3, 51.6% up); strongest faders DIA/XLE/XLI/KRE/MDY/SPY.
ExpressFade intraday weakness in GVA, BE, AMAT, ROAD, UCTT from ~10:00 ET toward the close.
FramingA small positive-expectancy mean-reversion, not a conviction long — tiny edge, tight risk.
InvalidationInvalidated if the gap keeps trending down through 10:30 on expanding volume.
Low convictionwarehouse:09
Event risk — crowded short float
TWST short float 28.4% (13 names >=20%)
High short-float names carry ~2.5x baseline odds of a +15% 10-day move (16.9% vs 6.8%), in BOTH directions.
ExpressTreat TWST, HIMS, MARA, APLD as elevated-variance event risk; size for a two-sided move, not a squeeze long.
FramingEvent risk, not a buy signal — the distribution is fat-tailed both ways.
InvalidationNot a trade until a catalyst and confirmation align.