StoQuant: AI-Powered Quantitative Stock Analysis

StoQuant is an institutional-grade quantitative stock analysis platform that scores 822 stocks daily across the S&P 500 and Russell 2000 using a 93-feature machine-learning pipeline. We help sophisticated retail investors find undervalued small-cap opportunities before Wall Street notices.

What makes StoQuant different

• Benjamin Graham intrinsic value: V = EPS × (8.5 + 2g), with a 30% margin-of-safety gate.
• 3-model ML ensemble: LightGBM + CatBoost + XGBoost stacked with a logistic-regression meta-learner and Venn-ABERS conformal calibration.
• Walk-forward proof: every alpha score is tracked against real forward returns. No cherry-picking, no hindsight.
• Black-Litterman optimizer: ML forecasts become investor views (Q vector); model confidence intervals map to the Ω uncertainty matrix.
• HMM regime detection: Hidden Markov Model with 3+ states (BIC-selected) classifies the market as low, normal, or high volatility.

Who it is for

Sophisticated self-directed investors, RIAs, and quant researchers who want transparent methodology and out-of-sample evidence — not a black-box newsletter.