ML Market Map — Japan Equity Clusters for 2026-08-28

A daily unsupervised machine-learning read of the Japan Equity market: 4,120 stocks grouped into 8 consensus clusters (KMeans + Gaussian-mixture + hierarchical, over robust-scaled PCA features) for 2026-08-28. 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

• low volatility (1y) · low idio vol — 1,631 names; mostly Industrials; drivers: rv 252 (−0.39), residual vol 252 d (−0.39), r2 252 d (+0.34)
• deep 1y drawdowns · deep drawdown · high downside vol — 907 names; mostly Industrials; drivers: pa mae 52w (−0.96), distance from high 252 (−0.75), max drawdown 252 (−0.73)
• high leverage · cheap (high B/P) · high cash yield — 440 names; mostly Industrials; drivers: leverage debt to mcap (+2.49), val bp z (+1.50), val cfop z (+1.38)
• strong 1y momentum · wide 1y upside range · high beta — 283 names; mostly Industrials; drivers: r2 252 d (+0.99), ret 252d (+0.81), pa mfe 52w (+0.70)
• wide 1y upside range · attention spent · high beta — 254 names; mostly Technology; drivers: pa mfe 52w (+2.61), pa sir recovered (+2.57), beta 252 d (+2.55)
• attention spent · high turnover · expensive (low E/P) — 212 names; mostly Industrials; drivers: pa sir recovered (+2.44), turnover to mcap (+2.36), vol of vol 63 (+2.25)
• rising earnings — 205 names; mostly Industrials; drivers: cfo growth (+4.15), ni growth (+1.04), pa sir infectious (−0.33)
• rising (3m) · high turnover · wide upside excursions (21d) — 188 names; mostly Communication Services; drivers: ret 63d (+2.49), ret 20d (+2.16), ret 126d (+1.95)

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=jp — clusters, names, sector mix, representative tickers, top features
• /api/ml/tickers?market=jp — per-ticker cluster, confidence, anomaly score, day-over-day migration, tenure
• /api/ml/pca?market=jp — 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.