ML Market Map — Taiwan Equity Clusters for 2026-08-31

A daily unsupervised machine-learning read of the Taiwan Equity market: 2,105 stocks grouped into 6 consensus clusters (KMeans + Gaussian-mixture + hierarchical, over robust-scaled PCA features) for 2026-08-31. 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 · high leverage — 606 names; mostly Consumer Cyclical; drivers: rv 252 (−0.61), residual vol 252 d (−0.60), leverage debt to mcap (+0.57)
attention spent · high turnover · high beta — 417 names; mostly Technology; drivers: r2 252 d (+0.76), pa sir recovered (+0.73), turnover to mcap (+0.68)
high leverage · falling earnings · deep pullbacks (21d) — 344 names; mostly Technology; drivers: leverage debt to mcap (+1.11), ni growth (−0.94), cfo growth (+0.73)
high growth · strong 1y momentum · high turnover — 321 names; mostly Technology; drivers: cfo growth (−1.16), revenue growth (+1.13), ret 252d (+1.12)
high turnover · attention spent · strong 1y momentum — 238 names; mostly Technology; drivers: turnover to mcap (+3.75), pa sir recovered (+3.61), ret 252d (+2.76)
low margin · expensive (low E/P) · deep pullbacks (21d) — 179 names; mostly Technology; drivers: profit margin (−3.27), val ep z (−2.19), pa mae 21d (−1.02)

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