find-untapped-thesislisted

# Find Untapped Thesis — NOK-Style 未爆发 Hunting ## Goal Find stocks that meet ALL of these criteria: 1. **未涨过头**: 1Y return < 50% (vs sector that's +100%+) 2. **估值便宜**: Forward P/E < 25 OR PEG < 1.2 3. **真 catalyst**: Concrete contracts, financials confirm thesis (not just narrative) 4. **机构未挤进**: Institutional ownership <30% (room for re-rating) 5. **Insider 不在大量卖**: Use insider_ratio.py to filter out distribution The classic pattern that captures multi-bagger reversal: low PE + low attention + new narrative tailwind + low institutional ownership + insider conviction. ## Why this matters **Crowded names = limited upside** (NVDA, AVGO, AMD already +100%+ YTD). **Untapped names = asymmetric R/R** because: - Estimates can still be revised up - Multiple expansion possible (PE 12 → 25 = +108%) - Institutional buying creates flow tailwind - Less covered by analysts = slower price discovery ## The 6-Step Workflow ### Step 1 — Define theme (user asks or you suggest) Common themes for 2026: - **AI Data Center power** (electricity, gas, nuclear) — EQT, AEP, ETR - **AI infrastructure 2nd derivative** (cooling, busbar) — NVT, MOD, ENS - **Memory recovery** (HBM, HDD) — MU, WDC - **Industrial gases for fabs** — APD, LIN - **Copper for AI buildout** — HBM, FCX, TECK - **Optical networking 2.0** — TEL, FN, CIEN - **EU defense / aerospace aluminum** — CSTM - **Quantum computing** (longer-term) — IONQ, RGTI, QBTS - **Robotics / industrial automation** — ROK, EMR ### Step 2 — Build ca