ML Market Map — Hong Kong Equity Clusters for 2026-07-16

A daily unsupervised machine-learning read of the Hong Kong Equity market: 2,419 stocks grouped into 7 consensus clusters (KMeans + Gaussian-mixture + hierarchical, over robust-scaled PCA features) for 2026-07-16. Descriptive, not predictive — there is no buy or sell signal. Research, not investment advice.

Markets: US Equity · China A-Share · Japan Equity · Hong Kong Equity · Taiwan Equity · Korea Equity · UK Equity · India Equity

Today's clusters

cheap (high E/P) · high margin · rising earnings — 837 names; mostly Industrials; drivers: r2 252 d (+0.49), val ep z (+0.46), profit margin (+0.45)
high downside vol · high volatility (3m) · high volatility (1y) — 568 names; mostly Industrials; drivers: vol of vol 63 (+0.87), downside vol 63 (+0.78), rv 63 (+0.70)
high turnover · heavily traded · high beta — 275 names; mostly Healthcare; drivers: turnover to mcap (+3.83), r2 252 d (+1.11), log turnover (+1.02)
expensive (low E/P) · high leverage · low margin — 267 names; mostly Real Estate; drivers: val ep z (−4.04), leverage debt to mcap (+2.56), profit margin (−1.56)
strong 1y momentum · high volatility (1y) · rising (3m) — 190 names; mostly Industrials; drivers: ret 252d (+3.06), ret 126d (+2.20), rv 252 (+1.21)
high leverage · low margin — 164 names; mostly Real Estate; drivers: leverage debt to mcap (+4.32), r2 252 d (+0.44), profit margin (−0.35)
rising (3m) · strong 1y momentum · high volatility (3m) — 118 names; mostly Industrials; drivers: ret 63d (+4.21), ret 126d (+4.18), ret 252d (+3.51)

Machine-readable data (free, read-only JSON)

The full map, per-ticker cluster assignments with confidence and anomaly scores, and PCA structure are published as open JSON for automated and AI-analyst consumption:

/api/ml/latest?market=hk — clusters, names, sector mix, representative tickers, top features
/api/ml/tickers?market=hk — per-ticker cluster, confidence, anomaly score, day-over-day migration, tenure
/api/ml/pca?market=hk — PCA components, explained variance, feature loadings

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.