PUBLIC EQUITY RESEARCH
SK HYNIX INC
South Korea · 000660
- Latest published edition
- 2026.07.19.2
- Report generated
- July 19, 2026
Reviewed research summary
Published research overview
The public summary is arranged into shorter paragraphs for faster scanning without changing its published wording.
This independent deep due diligence report on SK hynix (KRX: 000660) delivers a Hold / Wait for a Wider Margin of Safety rating with medium confidence, a 12-month probability-weighted target price of KRW 1,970,000, and expected upside of just 6.9% versus the July 16, 2026 KRX closing price of KRW 1,842,000.
The report applies a three-way stress test of HBM leadership durability, conventional memory cycle dynamics, and valuation assumptions, distinguishing corporate quality from stock pricing: SK hynix holds a market-leading 56.4% Q1 2026 HBM share, has initiated M15X wafer production for HBM4, and maintains strong liquidity after a USD 26.2bn U.S. offering, but the current share price already embeds expectations of sustained peak margins, uninterrupted leadership, and flawless execution.
The report notes Q1 2026 operating margin reached 72%, an extreme cyclical level, with net income inflated by KRW 9.9tn of Kioxia-related investment valuation gains and KRW 4.0tn of nonrecurring dividend income, creating a risk investors value peak cyclical earnings at a permanent multiple. The analysis decomposes SK hynix’s business into DRAM (77.3% of Q1 revenue, including HBM), NAND (22.0%, including Solidigm), and other smaller operations, highlighting that HBM revenue is not separately disclosed, so precise generation-level ASP, yield, and customer qualification data remain unverifiable, requiring a range-based approach to valuation.
A 10-year cycle review and 100-point cycle thermometer score current conditions at 76.4, indicating late-cycle expansion with strong pricing power but a narrowing forward margin of safety. The report builds a ground-up HBM demand model across conservative, base, and bull scenarios, with the base case projecting 119% of 2026 HBM revenue in 2027, while accounting for risks of customer multisourcing, Samsung and Micron competitive catch-up, and ASP normalization.
Capacity analysis tracks M15X, the first Yongin fab, P&T7 advanced packaging, and the Indiana packaging project, emphasizing that expansion adds both future supply and depreciation, creating a multiyear depreciation staircase rather than immediate earnings upside. Geopolitical risk analysis covers annual U.S. export licenses for Chinese fabs, CHIPS Act subsidy guardrails, and China NAND pricing constraints from the Intel NAND acquisition.
Financial analysis normalizes Q1 2026 net income to remove nonrecurring gains, builds an 8-quarter earnings path and 5-year financial model that preserves cyclical margin compression, and calculates valuation via four methods: DCF, mid-cycle P/E, EV/EBITDA, and P/B, yielding a weighted mid-cycle value of KRW 1,614,000, 12.4% below the current price. Reverse valuation shows the current share price requires terminal free cash flow 41% above the report’s base case to justify its level.
The report defines four weighted scenarios (Bull 20%, Base 45%, Bear 25%, Stress 10%) with corresponding 12-month targets, staged trading bands from KRW 950,000 to KRW 3,000,000, position limits for different risk profiles, and a 40-item monitoring dashboard to track metrics including pricing, market share, working capital, and project milestones. A 12-trigger Kill Thesis mandates immediate position freezes or reductions for events such as HBM4 qualification delays of more than two quarters, HBM share falling below 45% for two consecutive quarters, or nonrenewal of Chinese fab export licenses cutting output by more than 10%.
Red-team objections highlight risks of a low P/E trap at peak earnings, temporary HBM leadership, capital expansion sowing the next supply glut, overstated core earnings, undisclosed customer concentration, low-return Solidigm assets, nonlinear China policy risk, dilution from the recent offering, and insufficient upside for cyclical error. The report concludes an upgrade to Buy requires simultaneous confirmation of HBM4 cash profit generation, HBM4E qualification progress, working capital normalization, and project ROIC exceeding the cost of capital, while a single Kill Thesis trigger is sufficient for a downgrade.
The analysis adheres strictly to evidence discipline, labeling all claims as confirmed facts, third-party estimates, reasoned inferences, model calculations, or unverifiable data, and includes a full bibliography of primary sources to ensure reproducibility.
Report directory
Core modules included in this edition
- 01Research Scope, Evidence Discipline, and Reader's Guide
- 02Executive Summary and Investment Decision
- 03Business Mix, Revenue Quality, and Customer Concentration
- 04Ten-Year Cycle Review and Cycle Thermometer
- 05HBM Deep Dive: Evidence Ladder, Demand Model, and Supply Bottlenecks
- 06Competitive Landscape and Moat Stress Test
- 07Capacity, Capital Expenditures, and the Depreciation Cliff
- 08China, Export Controls, and Geopolitics
- 09Korean Governance, Shareholder Returns, and Foreign Exchange
- 10Financial Quality, Earnings Normalization, and Cash Conversion
- 11Valuation, Scenario Path, and Trade Execution
- 12Monitoring Dashboard: 40 Actionable Indicators
- 13Kill Thesis: 12 Triggers That Must Change the Conclusion
- 14Red-Team Review, Evidence Boundaries, and Investment-Committee Questions
- 15Investment Conclusion
- 16Principal Sources and Reproducible Index
FREE-POOL DOSSIER PREVIEW
Read the opening section of this investment dossier
This server-generated preview contains the first 12 of 43 pages and never exceeds 30% of the report.