Alright—let’s do this the Munger way: cut the fluff, anchor on unit-economics and moat, and don’t overpay for cyclical euphoria.
Recommended buy price (Tokyo: 6857): ¥12,000–13,000 for steady accumulation; strong buy zone ~¥10,500 (≈ ~35× forward EPS).
Current price (Sep 26, 2025): ¥14,770 (delayed ~15m).
Advantest — GPT Investor Master Checklist
1) Company Overview
1.1 What do they do? Global leader in semiconductor test equipment—SoC and memory testers plus handlers, interfaces, services/system-level test.
1.2 Sector/sub-sector: Semiconductors → Automated Test Equipment (ATE).
1.3 TAM: Semi test equipment market ~$7–8B in 2025, mid-single to high-single-digit CAGR to ~$11–13B by early 2030s. (Multiple sources; methodologies differ.)
1.4 Competitors & share: Duopoly dynamics with Teradyne; together ~80% share; Advantest share rising in SoC testers.
1.5 Concentration: Customers concentrated in leading-edge foundry/OSAT and GPU/AI ecosystems (TSMC/NVIDIA/AMD via foundry routes); cyclical but recurring with each node/test step. (Disclosure & analyst synthesis.)
2) Quick Screen
2.1 Debt-to-equity: ~16% (mrq, LSEG/Yahoo calc). Net cash position substantial.
2.2 Address: Shin Marunouchi Center Bldg., 1-6-2 Marunouchi, Chiyoda-ku, Tokyo 100-0005, Japan.
2.3/3 Scandal/Fraud (company/promoters): No material fraud controversies noted in recent filings/news. (Ongoing check advised.)
2.5 Debtor days: Not disclosed as “days” on Reuters page; use AR/quarterlies for precise calc (trend stable given blue-chip mix).
2.6 Market cap: ~¥11.5T at current price.
2.7 Cash & equivalents: ~¥273.4B (Jun 30, 2025).
2.8 FCF: Positive through cycle; FY25 Q1 operating cash flow ¥46.9B; full-year typically tracks earnings.
2.9 FX exposure: Material—yen moves affect results; management flags FX volatility risk.
3) Shareholder Alignment
3.2/3.3 Promoters & ownership: Japanese public company with professional management/board; no promoter pledges typical of India. (See IR & Reuters leadership.)
3.5 Openness: Strong disclosure cadence—quarterly results, slides, Q&A, webcasts.
3.9 Dividend policy: Modest yield (~0.26%)—growth + reinvestment focus.
4) Performance
4.1 Price: ¥14,770 (today).
4.2 PE (forward): ~46× at current level (LSEG).
4.3 Sales trend: FY2024 (Mar-end 2025) revenue ~¥780B; FY2025 guide ~¥755B (pre-raise), with July raise on AI demand.
4.4 PAT: FY2025 (guide updated) op. profit raised to ¥300B as of Jul 29, 2025 (implies strong net income uplift).
4.5 EPS trend: Rising with AI tester demand; (use FY result + share-count for exact calc in a model).
4.6 Margin stability: Operating margins expanding with AI mix/node transitions; management raised op. margin outlook.
4.7 Segment drivers: SoC testers + services/system-level test; memory test smaller but cyclical.
5) Efficiency
5.1 OPM: Expanding YoY on mix; see quarterly slides for exact %.
5.2 ROCE / 5.3 ROE: High through cycle; ROE ~28% (TTM per LSEG page).
5.5 Debtor days / 5.6 Asset turns / 5.7 WC cycle / 5.8 Inventory days: Healthy, aided by prepay/long-lead dynamics; inventories ~¥209B at 6/30/25.
6) Financial Risk
6.1 Debt: Low absolute debt; net cash. D/E ~16% (mrq).
6.3 Future debt: No signal of large debt plans; focus on organic capex and R&D.
6.4 Credit rating: (Not prominently disclosed on IR; Japan issuers may use JCR/R&I—treat as N/A here.)
6.5/6.6 Contingent/regulatory: Standard export-control and China/US trade-policy sensitivity as with peers. (Sector context.)
7) Volume & Liquidity
7.1 Liquidity: Strong in Tokyo; ~3.5M shares traded today.
7.5 Volatility: Elevated vs TOPIX given AI tester cycle; beta ~1.55 (indicative).
8) Valuations
8.1/8.9 Multiples: Forward P/E ~46×; P/B ~19×; rich vs 5-yr averages (AI test cycle premium).
8.2 GARP sense: At ~46×, you want ≥20–25% durable EPS CAGR or you wait for better entry. (Rule-of-thumb framework.)
8.8 EV/EBITDA: High-teens to low-20s on FY26 street—fair for a quasi-duopoly, but offers little MOS at peaks. (Inference from forward PE/op. margin.)
9) Growth Components
9.5 Capex/R&D: Persistent reinvestment to maintain platform lead (V93000, SLT, analytics).
9.6 Moat: Scale + installed base + software/tooling + customer lock-in; duopoly economics.
9.9 R&D spend: Consistently high by ATE standards (see Investors’ Guide/AR).
9.10–9.12 Expansion/ESG/Succession: Global footprint; sustainability reporting; professionalized leadership bench.
Why my buy zones? (Buffett-ish sanity check)
Business quality: Near-duopoly with Teradyne, entrenched at cutting edge. ✔️
Cycle risk: AI tester upcycle can mean over-earning—multiple + margins both elevated. ❗
MOS discipline: I want ≤ ~40–45× forward for steady nibbling (¥12k–13k), and ≤ ~35× for size-up (~¥10.5k). This bakes in some cycle giveback while paying up for quality. Cross-check with your own EPS bridge (FY26/27).
New frameworks to pressure-test before buying
- Duopoly Stress Test: Model market-share slips of 200–300 bps vs TER and see EPS sensitivity. (Use AR unit volumes + mix.)
- Node/Packaging Exposure Map: Attribute revenues to HBM/GPU wafer sort, SLT, and memory test—score cyclicality vs AI substrate.
- FX Shock Table: ±10% USD/JPY move on operating income (use company FX sensitivities in results slides).
- Installed-Base Flywheel: Track service attach rate & software dollars per installed tester—this is your annuity moat metric.
- Through-Cycle P/E Bands: Build 10-year band (bear/mid/bull) and only add when price ≤ mid-band unless growth proves step-function.
Horizons 
Leave a comment