A daily unsupervised machine-learning read of the US Equity market: 4,802 stocks grouped into 10 consensus clusters (KMeans + Gaussian-mixture + hierarchical, over robust-scaled PCA features) for 2026-08-27. Descriptive, not predictive — there is no buy or sell signal. Research, not investment advice.
high leverage · high margin · low volatility (3m) — 1,820 names; mostly Financial Services; drivers: leverage debt to mcap (+0.52), profit margin (+0.41), rv 63 (−0.39)
expensive (low E/P) · low cash yield · falling earnings — 575 names; mostly Healthcare; drivers: val ep z (−2.52), val cfop z (−1.89), ni growth (−1.43)
wide upside excursions (21d) · strong 1y momentum · wide 1y upside range — 550 names; mostly Healthcare; drivers: ret 20d (+1.10), pa mfe 21d (+1.03), ret 252d (+1.03)
deep pullbacks (21d) · high turnover · attention spent — 493 names; mostly Healthcare; drivers: pa mae 21d (−1.29), turnover to mcap (+0.82), pa sir recovered (+0.80)
rising earnings · expensive (low E/P) · high growth — 433 names; mostly Financial Services; drivers: cfo growth (+2.23), ni growth (+1.01), val ep z (−0.65)
attention spent · high turnover · wide upside excursions (21d) — 223 names; mostly Technology; drivers: pa sir recovered (+3.11), turnover to mcap (+2.94), pa mfe 21d (+2.11)
high leverage · expensive (low E/P) · high cash yield — 213 names; mostly Consumer Cyclical; drivers: leverage debt to mcap (+3.60), val ep z (−2.59), val cfop z (+2.58)
high downside vol · high volatility (3m) · attention spent — 212 names; mostly Healthcare; drivers: vol of vol 63 (+3.85), downside vol 63 (+3.49), rv 63 (+3.11)
cheap (high B/P) · high cash yield · cheap (high E/P) — 170 names; mostly Financial Services; drivers: val bp z (+4.68), val cfop z (+4.63), val ep z (+4.58)
rising (3m) · high volatility (3m) · strong 1y momentum — 113 names; mostly Healthcare; drivers: vol of vol 63 (+4.08), ret 63d (+3.79), ret 126d (+3.33)
Machine-readable data (free, read-only JSON)
The full map, per-ticker cluster assignments with confidence and anomaly scores, PCA structure, and within-cluster cointegrated relative-value pairs are published as open JSON for automated and AI-analyst consumption:
/api/ml/latest — clusters, names, sector mix, representative tickers, top features
/api/ml/pairs — within-cluster cointegrated pairs (Engle-Granger, OU half-life, Kalman z-score; US only)
Descriptive market-structure research only. Unsupervised clustering finds structure, not direction; a tight cluster or an anomaly is a starting point for research, never a trade signal.