photonicsense · peer comparison
Why SK Hynix is the anomaly
A side-by-side read on SK Hynix versus the three companies that define the AI memory/packaging stack: Micron (the closest memory peer with a US listing), Samsung (the next-largest HBM competitor), and TSMC (the packaging bottleneck). At PEG 0.09 vs peers at 0.35-1.5+, the anomaly isn't subtle.
We own
SK Hynix
000660.KS · HY9H.F
Fwd PE5.85x
PEG0.09
Op Margin58.4%
Rev Growth+66% YoY
HBM Share62%
US ListedPending Q2/Q3
Peer
Micron
MU · NASDAQ
Fwd PE~8x
PEG~0.35
Op Margin~28%
Rev Growth+148% YoY
HBM Share~15%
US ListedYes
Peer
Samsung
005930.KS · SMSN.IL
Fwd PE~7x
PEG0.18
Op Margin~15%
Rev Growth~20%
HBM Share~23%
US ListedGDR only
Peer
TSMC
TSM · NYSE
Fwd PE~22x
PEG~0.6
Op Margin~45%
Rev Growth+34% YoY
HBM Share— packager
US ListedYes · 30% premium
Forward PE — The Discount
SK Hynix at 5.85x vs Micron 8x vs TSMC 22x. The gap is historical and structural — ADR closes it.
PEG Ratio — Growth-Adjusted Value
Lower is cheaper for the growth. SK Hynix at 0.09 is extraordinary. No mature semi leader has traded here recently.
Full Comparison Matrix
Every metric that matters for the thesis, side by side
| Metric | SK Hynix | Micron | Samsung | TSMC |
| Forward PE | 5.85x | ~8x | ~7x | ~22x |
| Trailing PE | 17.2x | ~19x | ~17x | ~30x |
| PEG Ratio | 0.09 | ~0.35 | ~0.18 | ~0.6 |
| EV / Revenue | 7.3x | ~5x | ~1.8x | ~11x |
| Price / Book | ~6x | ~4x | ~1.5x | ~8x |
| — | Growth & Profitability ↓ |
| Revenue (TTM) | ₩97.2T | ~$40B | ~₩340T | ~$95B |
| Revenue Growth | +66% YoY | +148% YoY | ~+20% | +34% YoY |
| Operating Margin | 58.4% | ~28% | ~15% | ~45% |
| Net Margin | 44.2% | ~20% | ~10% | ~42% |
| ROE | 44.1% | ~18% | ~10% | ~32% |
| Net Cash | ₩35T (~$25B) | Net debt | ~₩92T | Large |
| — | Memory / HBM Position ↓ |
| HBM Revenue Share | 62% | 15% | 23% | — CoWoS packager |
| HBM4 Qualified | Yes (Sep '25) | Q2 '26 target | Yes (Feb '26) | Base die partner |
| Nvidia HBM4 Share | ~70% | ~0% (yr 1) | ~30% | Exclusive CoWoS |
| EUV Strategy | All-in · memory | Avoiding | Split fdn + mem | Foundry leader |
| — | Catalysts & Risk ↓ |
| US Listing Status | ADR Q2/Q3 2026 | Listed | GDR only (LSE) | Listed ADR (+30%) |
| Analyst Consensus | 36B / 1S | Strong Buy | Buy | Strong Buy |
| Price Target | ₩1.39M (+20%) | +15% | +12% | +8% |
| Biggest Risk | Korea geopolitics | HBM4 qual delay | Conglomerate | Valuation |
Investment Scorecard
Subjective scoring across 6 dimensions. 5 stars = best in class.
SK Hynix
Valuation★★★★★
Growth★★★★☆
Profitability★★★★★
Moat★★★★★
Catalyst★★★★★
Risk★★★☆☆
TOTAL27/30
Micron
Valuation★★★★☆
Growth★★★★★
Profitability★★★☆☆
Moat★★☆☆☆
Catalyst★★★☆☆
Risk★★★☆☆
TOTAL20/30
Samsung
Valuation★★★★★
Growth★★★☆☆
Profitability★★★☆☆
Moat★★★☆☆
Catalyst★★☆☆☆
Risk★★★★☆
TOTAL20/30
TSMC
Valuation★★☆☆☆
Growth★★★★☆
Profitability★★★★★
Moat★★★★★
Catalyst★★☆☆☆
Risk★★☆☆☆
TOTAL20/30
The takeaway
SK Hynix scores 27/30. Micron, Samsung, and TSMC all score 20/30. The difference is the simultaneous combination of (a) lowest PE / PEG among meaningful peers, (b) highest operating margins, (c) highest HBM market share, and (d) a discrete, time-bound catalyst (ADR). Micron has growth but weaker moat and still sub-Hynix margins. Samsung has the cheapest valuation but carries conglomerate dilution and hasn't taken HBM share at scale. TSMC is the "picks-and-shovels" monopoly but you pay 22x for the privilege.
The trade is: own the cheap stock with the catalyst, not the expensive stock with the moat.