ML Market Map — Japan Equity Clusters for 2026-07-16

A daily unsupervised machine-learning read of the Japan Equity market: 4,111 stocks grouped into 10 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

• broad market · slight low volatility (1y) — 1,487 names; mostly Industrials; drivers: rv 252 (−0.32), residual vol 252 d (−0.31), profit margin (+0.31)
• high turnover · deep drawdown · high downside vol — 435 names; mostly Communication Services; drivers: turnover to mcap (+1.45), vol of vol 63 (+1.29), max drawdown 252 (−1.29)
• poor risk-adj return · deep drawdown · falling (3m) — 427 names; mostly Industrials; drivers: distance from high 252 (−0.46), ret 126d (−0.40), rolling sharpe 63 (−0.39)
• high leverage · high cash yield · cheap (high B/P) — 373 names; mostly Industrials; drivers: leverage debt to mcap (+2.98), val cfop z (+1.41), val bp z (+1.31)
• rising (3m) · strong 1y momentum · high risk-adj return — 340 names; mostly Industrials; drivers: ret 63d (+1.47), ret 126d (+1.32), ret 20d (+1.30)
• rising earnings · cheap (high B/P) · high leverage — 308 names; mostly Industrials; drivers: cfo growth (+3.37), ni growth (+1.23), val bp z (+0.42)
• high beta · strong 1y momentum · rising (3m) — 296 names; mostly Unknown; drivers: r2 252 d (+2.66), beta 252 d (+1.29), ret 252d (+1.24)
• expensive (low E/P) · high turnover · high volatility (1y) — 183 names; mostly Communication Services; drivers: val ep z (−2.30), turnover to mcap (+2.26), vol of vol 63 (+2.22)
• strong 1y momentum · high turnover · high beta — 138 names; mostly Technology; drivers: ret 252d (+3.50), ret 126d (+3.38), turnover to mcap (+3.29)
• falling earnings · expensive (low E/P) · low margin — 124 names; mostly Industrials; drivers: ni growth (−2.82), val ep z (−2.19), profit margin (−1.55)

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.