Skip to main content

SK Hynix - Comprehensive Investment Analysis 2026

  • Analysis Date: August 3, 2026
  • Exchange: KRX (primary) / NASDAQ ADR "SKHY" (2026 US listing)
  • Sector: Semiconductors - Memory (DRAM, NAND, HBM)
  • Market Cap: KRW 1,142 trillion ($792B), world's ~18th most valuable company

Executive Summary

Fundamental Score: 8.0/10 ⭐⭐⭐⭐

Investment Recommendation: Buy (Accumulate on Weakness)

Conviction Level: Medium-High

Target Price: KRW 1,800,000-2,200,000 (15-40% upside) / ADR $170-200 — 12-month horizon

Key Thesis: SK Hynix is the dominant supplier of High-Bandwidth Memory (HBM) for AI accelerators, holding ~63% of the HBM market and an estimated ~70% share of upcoming HBM4 volumes for Nvidia's Rubin platform. Q2 2026 results were record-breaking (revenue +257% YoY, operating margin 76%) but still missed elevated analyst expectations, triggering an ~10% single-day selloff — a reminder that a near-perfect execution is now priced in. The balance sheet has swung to a net cash position (net debt/equity -26%), giving the company unusual flexibility to fund KRW 40T+ capex for HBM4 and the new Yongin fab without financial strain. The core debate is not whether the AI memory supercycle is real, but how much of it is already in the KRW 1.14 quadrillion market cap.

Business Overview

Company Profile:

SK Hynix Inc, headquartered in Icheon, South Korea, is the world's second-largest memory chipmaker (after Samsung Electronics) and the clear leader in High-Bandwidth Memory (HBM) used in AI accelerators. The company traces its roots to 1949 (as Hyundai Electronics) and was acquired into the SK Group in 2012. It listed a US ADR ("SKHY") in 2026 alongside its long-standing KRX listing (000660).

Business Model:

  • Revenue Streams: DRAM (~65-70% of revenue), NAND flash (~25-30%), other (NOR, foundry services)
  • Key Products: HBM3/HBM3E/HBM4 for AI GPUs (Nvidia H100/H200/Rubin), DDR5/LPDDR for servers and mobile, enterprise SSDs (eSSD)
  • Distribution: Direct supply agreements with hyperscalers (Nvidia, AMD, Google, Microsoft, Amazon) and OEMs

Market Position:

  • HBM: #1 globally, ~63.2% revenue share (2025); UBS estimates ~70% share of HBM4 in 2026
  • DRAM: #1 globally by revenue in FY2025 (34.8% share), slipped to #2 in Q1 2026 (29.1%) as Samsung regained share
  • NAND: #2 globally (20.9% share, 2025), strength concentrated in eSSD
  • Key Competitors: Samsung Electronics (memory division), Micron Technology, Kioxia/Western Digital (NAND)

Competitive Moat:

  • First-mover advantage in HBM — first to mass-produce multiple HBM generations (HBM3, HBM3E, HBM4)
  • Oligopoly structure: DRAM/HBM production is effectively a 3-player market (SK Hynix, Samsung, Micron), high capital and process barriers to entry
  • Deep, multi-year supply agreements with Nvidia tying HBM roadmap to GPU roadmap (Blackwell → Rubin)

Management Quality:

  • Operates under SK Group's semiconductor strategy, governed via SK Square (spun off from SK Telecom in 2021 specifically to house the group's chip bet)
  • Track record of successfully navigating the 2023 memory downturn (net loss year) into a 2024-2026 record-profit recovery

Corporate Governance:

  • Largest Shareholder: SK Square, ~20.5% (itself 100%+ controlled via SK Inc., 17.9%-owned by SK Group Chairman Chey Tae-won)
  • National Pension Service (Korea): ~8.1%
  • BlackRock Fund Advisors: ~5.1% (largest disclosed foreign holder)
  • Total foreign institutional ownership: ~54% (BlackRock, Vanguard, GIC, and others below disclosure thresholds)
  • No pledge disclosure applicable (Korean market does not use the promoter-pledge framework common to Indian markets)

Financial Analysis

Annual Performance (KRW Trillion)

MetricFY2022FY2023FY2024FY2025Trend
Revenue44.6232.7766.1997.15Sharp recovery + AI-driven acceleration
Operating ProfitN/ALoss23.4747.21Margin nearly doubled YoY in FY2025
OPM (%)N/ANegative~35%~49%Structural HBM mix-shift improving margins
Net ProfitN/ALoss19.8042.95Net margin ~44% in FY2025

FY2023 was a loss-making trough (memory downcycle); FY2024-2025 mark a sharp V-shaped recovery driven by the AI memory supercycle. (Data not available: precise FY2022/FY2023 operating and net profit figures — flagged as gap.)

Quarterly Performance (Recent Quarters, KRW Trillion)

QuarterQ4 FY2025Q1 FY2026Q2 FY2026
Revenue32.8052.5879.32
Operating Profit19.2037.6160.54
OPM (%)~58.5%~71.5%~76.3%

Q2 2026 revenue grew +50.9% QoQ and +256.8% YoY; operating profit grew +61.0% QoQ and +557.2% YoY — a genuinely record quarter. However, it still missed analyst consensus (LSEG SmartEstimates expected ~KRW 84T revenue / KRW 64T operating profit), and the stock fell as much as -15% intraday, closing -9.6% lower.

Margin Analysis

MarginFY2024FY2025Q2 FY2026Trend
OPM~35%~49%~76%Structurally expanding on HBM mix-shift
Net Margin~30%~44%Data not available (quarterly net-profit figures from public sources were inconsistent — flagged, not used)Expanding

Cash Flow Quality

MetricValueAssessment
TTM Free Cash Flow~$28.1B (per GuruFocus, TTM to Mar 2026)Strong, though compressed by aggressive capex ramp
FY2026E CapExKRW 40T+ (upper end of guided range vs KRW 30.2T in FY2025)Funding HBM4 ramp + Yongin Phase 1 fab (opening early 2027)
FY2026E Operating Cash Flow~KRW 50T (estimated)Implies theoretical FCF of ~KRW 15T after the capex step-up

Cash Flow Rating: 7.5/10 — cash generation is strong but capex intensity is rising sharply, compressing near-term FCF even as profitability improves.

Balance Sheet Strength

MetricQ2 FY2026Analysis
Cash & EquivalentsKRW 87.96T (up from KRW 54.33T prior quarter)Rapidly building war chest
Total DebtKRW 18.59T (down from KRW 19.32T)Actively deleveraging despite capex ramp
Net Cash PositionKRW 69.4TStrongly net-cash
Debt-to-Equity7% (down from 12% prior quarter, 25% a year earlier)Excellent, rapidly improving
Net Debt-to-Equity-26%Net cash exceeds debt by a wide margin

Balance Sheet Rating: 9.0/10 — among the strongest balance sheets of any capital-intensive semiconductor manufacturer globally.

Shareholding Pattern

CategoryHoldingNotes
SK Square (largest shareholder)~20.5%Strategic control vehicle, spun off from SK Telecom (2021)
National Pension Service (Korea)~8.1%Domestic institutional anchor
BlackRock Fund Advisors~5.1%Largest disclosed foreign holder
Other foreign institutions (Vanguard, GIC, etc.)~54% total foreign ownership (incl. above)High foreign institutional confidence
Public/Other~66% (residual, below 5% disclosure thresholds)Widely held

Shareholding Rating: 7.5/10 — no promoter-pledge risk (not applicable to Korean listing structure); high foreign institutional participation (~54%) signals global investor confidence, though SK Square's ~20.5% stake means ultimate control sits with SK Group Chairman Chey Tae-won despite his holding no direct SK Hynix stock.

Valuation

MetricSK Hynix (KRX:000660)Micron (NASDAQ:MU)*Assessment
Current PriceKRW 1,567,000$946.96
Market CapKRW 1,142T ($792B)~$1.07TSK Hynix is ~26% smaller by market cap
PE Ratio (Trailing)6.87x (stockanalysis.com)~45x TTM (per prior KB analysis)SK Hynix trailing PE looks far cheaper, but two sources disagree materially (companiesmarketcap.com quotes TTM PE of 16.5x) — flagged as a data-quality caveat, not resolved
PB Ratio4.35xData not fetched
PS Ratio6.04x~16.9x (per prior KB analysis)SK Hynix screens cheaper on sales too
Dividend Yield0.17%~0.06%Both minimal
Beta (5Y)2.322.17Both high-volatility, cyclical names

*Micron figures sourced from the existing Micron Technology comprehensive analysis (data as of June 9, 2026) — not re-verified today; shown for directional comparison only.

Valuation Verdict: FAIR TO CHEAP, with a caveat. On a trailing basis SK Hynix looks inexpensive relative to its own growth and to Micron, but (a) trailing PE is distorted upward by an extraordinary TTM that already embeds the AI-boom quarters, and (b) two independent sources disagree on the trailing PE (6.87x vs 16.5x) — likely a stale-data or EPS-calculation-basis issue on one side. Given the Q2 2026 earnings miss and post-earnings selloff, the market is actively re-rating how much of the AI memory supercycle is durable versus already realized.

Valuation Rating: 7.0/10

Fundamental Score: 8.0/10

CriteriaScoreRationale
Business Quality9/10Dominant HBM moat (63% share), oligopoly DRAM/NAND structure, deep Nvidia ties
Financial Health8/10Explosive growth, 76% quarterly OPM, but FCF compressed by capex ramp
Shareholding Pattern7.5/10No pledge risk, high foreign institutional ownership, clear SK Group control
Valuation7/10Screens cheap on trailing multiples but data conflicts and post-boom re-rating risk exist
Growth Prospects9/10HBM revenue forecast to grow from 31B(2025)to31B (2025) to 69B (2027), ~49.5% CAGR
Risk Management8/10Net-cash balance sheet, but single-customer (Nvidia) and cyclical-industry concentration

Composite Score: 8.0/10

Technical Summary

Overall Signal: Neutral-to-Bullish (mixed near-term, constructive medium-term)

Short-Term (1 month): Neutral — daily technical signal is neutral post-earnings selloff

Medium-Term (3-6 months): Bullish — 1-week and 1-month ratings both trend Buy; KRX-listed shares show 9 buy vs 3 sell signals across moving averages (the US ADR shows a weaker mixed picture, 4 buy vs 8 sell)

Action: Accumulate on Weakness

Price Action & Trend

Current Price: KRW 1,567,000 (Aug 3, 2026) / ADR $143.73

All-Time High: KRW ~2,987,000 (June 25, 2026) — current price is ~47% below that peak, reflecting a sharp de-rating after the Q2 earnings miss and broader profit-taking in AI-memory names

52-Week Change: +507% (extraordinary, reflects the AI memory supercycle re-rating over the past year)

ADR 52-Week Range: 124.80(low)124.80 (low) - 194.80 (high); currently trading near the lower-middle of that range at $143.73

Trend Assessment: The stock is in a powerful multi-quarter uptrend that has recently corrected sharply from an all-time high following an earnings-day selloff (-9.6% to -15% intraday on Jul 29, 2026 post-Q2-results). Divergence between KRX-listed and ADR technical signals suggests the correction has been more pronounced in USD/ADR terms than in local KRW terms.

Momentum & Support/Resistance

LevelPriceSignificance
All-Time High (resistance)KRW ~2,987,000 / ADR ~$194.80June 2026 peak, unlikely near-term retest
6-Month Resistance (ADR)~$193.92Matches 52-week high area
Current PriceKRW 1,567,000 / ADR $143.73
6-Month Support (ADR)~$151.16Recently broken to the downside post-earnings
ADR 52-Week Low$124.80Major downside support if selloff extends

Technical Risk Level: HIGH — Beta 2.32, stock has already corrected ~47% from its all-time high, and moving-average signals disagree between the KRX line and the US ADR, indicating unresolved directional consensus in the near term.

Fundamental vs Technical Alignment

DimensionSignalNotes
FundamentalBullishScore 8.0/10 — dominant HBM franchise, net-cash balance sheet, explosive (if slightly-missed) growth
TechnicalNeutral-to-BullishSharp post-earnings correction from ATH, but medium-term moving averages still favor buyers on the KRX line
AlignmentPartially AlignedFundamentals remain intact; the correction looks like a valuation reset after an "expectations exceeded reality" earnings reaction rather than a change in the underlying thesis
RecommendationBuy (Accumulate on Weakness)Combined view favors scaling in on further pullbacks toward the $124-151 ADR / KRW ~1,200,000-1,350,000 zone rather than chasing strength

Combined Conviction: Medium-High — the business case (HBM dominance, net-cash balance sheet, HBM4 ramp) is stronger than the technical picture, which is still digesting the post-earnings selloff. This favors a scaled entry rather than a single lump-sum purchase.

Investment Thesis

Bull Case (Target: KRW ~2,200,000 / ADR ~$200 — retest of all-time high)

  1. HBM4 ramp for Nvidia's Rubin platform (H2 2026) drives further share gains toward UBS's projected ~70% HBM4 share
  2. HBM revenue triples from 31B(2025)toaforecast31B (2025) to a forecast 69B (2027), a 49.5% CAGR, with SK Hynix capturing the majority
  3. Net-cash balance sheet allows continued capex funding (Yongin Phase 1 fab, opening early 2027) without dilution or leverage risk

Bear Case (Target: KRW ~1,000,000 / ADR ~$100 — further mean-reversion from ATH)

  1. AI capex growth among hyperscalers slows or plateaus, reducing HBM demand growth from "supercycle" to merely "strong"
  2. Samsung Electronics re-takes DRAM share aggressively (already #1 again in Q1 2026 at 29.1% vs SK Hynix), compressing pricing power
  3. Memory is a historically brutal cyclical industry — the FY2023 net-loss year is a reminder that oversupply corrections can be severe and fast

Base Case (Target: KRW ~1,800,000 / ADR ~$170)

Continued execution on HBM4 ramp and DRAM/NAND pricing strength through 2026-2027, with the stock recovering a meaningful portion of the post-Q2-earnings correction as the HBM4 ramp proves out, but without fully retesting the June 2026 all-time high given the market's now-heightened bar for "beating" already-extreme expectations.

Expected Return Distribution:

  • Bull Case (25% probability): +40% upside — KRW 2,200,000 / ADR ~$200
  • Base Case (50% probability): +15% return — KRW 1,800,000 / ADR ~$170
  • Bear Case (25% probability): -36% downside — KRW 1,000,000 / ADR ~$100

Risk Assessment

RiskProbabilityImpactDetails
AI capex slowdownMediumHighHyperscaler capex growth deceleration would directly compress HBM demand growth
Memory cyclicalityMediumHighFY2023 net-loss year proves the industry can swing from record profit to loss within 12-18 months
Customer concentration (Nvidia)MediumMedium-HighHBM revenue is heavily tied to Nvidia's GPU roadmap and order cadence
Samsung share recaptureMediumMediumSamsung retook #1 DRAM position in Q1 2026 (29.1% vs SK Hynix's declining share), pressuring pricing
Valuation data inconsistencyLow-MediumMediumConflicting trailing PE figures (6.87x vs 16.5x) across sources warrant independent verification before sizing a position
Geopolitical/export-control riskLow-MediumMediumKorea-China-US semiconductor trade tensions could affect equipment access or China sales

Overall Risk Rating: MEDIUM-HIGH (7/10) — strong fundamentals partially offset by extreme historical cyclicality and a name still digesting a sharp post-earnings correction.

Catalysts

Near-term (0-3 months):

  • HBM4 shipment ramp progress updates through H2 2026
  • Continued Nvidia Rubin platform rollout commentary

Medium-term (3-12 months):

  • Yongin Phase 1 cleanroom opening (early 2027) and associated capacity commentary
  • Q3/Q4 2026 earnings — market will be watching whether SK Hynix can beat (not just meet) elevated expectations after the Q2 miss

Long-term (1-3 years):

  • HBM market growth from 31B(2025)toaforecast31B (2025) to a forecast 69B (2027)
  • Potential further DRAM/NAND share shifts between SK Hynix, Samsung, and Micron as the three-player oligopoly reallocates the AI memory pie

Key Monitoring Parameters

Quarterly:

  1. HBM revenue mix and share trend (watch for erosion below ~60%)
  2. Operating margin trend (watch for reversal from the current ~76% peak, which is historically unsustainable)
  3. DRAM market share rank vs Samsung (SK Hynix slipped to #2 in Q1 2026 — watch whether this reverses or persists)

Trigger Events:

  • Buy more: Pullback toward ADR $124-151 / KRW 1,200,000-1,350,000 with HBM4 ramp still on track
  • 🔻 Reduce/Exit: Two consecutive quarters of HBM share loss to Samsung/Micron, or signs of a broader AI-capex pullback among hyperscalers

Conclusion

SK Hynix sits at the center of the AI memory supercycle with a genuinely dominant HBM franchise, a fortress net-cash balance sheet, and margins (76% operating in Q2 2026) that would have been unthinkable for a memory maker just two years ago. The Q2 2026 "record beat that still missed" and the resulting ~10% single-day selloff illustrate how much perfection is priced into the stock after a +507% 52-week run — the correction from the June 2026 all-time high looks more like an expectations reset than a thesis break. Given the conflicting valuation data points found across sources (trailing PE quoted as both 6.87x and 16.5x), position sizing should be conservative until this is independently verified, and entries are better scaled in on weakness than chased at strength.

Investment Recommendation: Buy (Accumulate on Weakness) Suitable For: Growth/momentum investors comfortable with high-beta (2.3+) cyclical semiconductor exposure and Korean-market/ADR liquidity considerations Portfolio Allocation: 3-6% of portfolio (high-conviction but volatile; size similar to or slightly below Micron given added valuation-data uncertainty)

Disclaimer

This analysis is for educational and informational purposes only and does not constitute investment advice. Conduct your own due diligence and consult a qualified financial advisor before making investment decisions.

Data Sources

  • stockanalysis.com — Valuation statistics, PE/PB/PS, EPS, shares outstanding (Accessed: August 3, 2026)
  • companiesmarketcap.com — Historical PE trend, market cap (Accessed: August 3, 2026)
  • SK hynix Newsroom (news.skhynix.com) — Q1/Q2 2026 and FY2025 financial results (Accessed: August 3, 2026)
  • CNBC — Q2 2026 earnings reaction, capex guidance (Accessed: August 3, 2026)
  • Investing.com / TradingKey / MarketScreener — Technical indicators, analyst price targets (Accessed: August 3, 2026)
  • GuruFocus — TTM free cash flow (Accessed: August 3, 2026)

Data Timestamp: August 3, 2026 (KRX Price: KRW 1,567,000; ADR Price: 143.73;MarketCap: KRW1,142T/ 143.73; Market Cap: ~KRW 1,142T / ~792B)

Note on data quality: Trailing PE ratio differs materially between sources (6.87x per stockanalysis.com vs 16.5x per companiesmarketcap.com). Quarterly net-profit figures for Q2 2026 also showed an inconsistency in one source (a figure exceeding operating profit, which is not plausible without a large one-off gain) and was excluded rather than reported as fact. Treat headline valuation multiples here as directional, not precise, until cross-verified against a primary filing (SK hynix's IR disclosures or DART/SEC filings).

Next Update Recommended: After Q3 2026 earnings (expected late October 2026)