EWY Part 2: SK Hynix ₩1M, Samsung ₩214K, KOSPI 6000 | EWY第二部分：海力士₩100万，三星₩21.4万，KOSPI 6000
February 26, 2026Part 2 Update — February 26, 2026
Original thesis published February 23, 2026. EWY closed at $139.30. Today: $149.62 (+7.4%).
Since our original memo, the KOSPI broke 6,000 for the first time in history. SK Hynix hit ₩1,000,000 — the first Korean stock ever to reach seven figures. Samsung punched through ₩200,000 on Feb 24 and trades at ₩214,500 today. EWY is up 24% month-to-date and +7.4% since Part 1.
The volatility thesis is playing out. Here's the updated picture.
Price Action Since Part 1
| Ticker | Part 1 (Feb 23) | Today (Feb 26) | Move |
|---|---|---|---|
| EWY | $139.30 | $149.62 | +7.4% |
| Samsung (005930.KS) | ₩193,000 | ₩214,500 | +11.1% |
| SK Hynix (000660.KS) | ₩949,248 | ₩1,018,000 | +7.2% |
| KOSPI | ~5,400 | 6,000+ | +11.1% |
EWY monthly trajectory (Feb 2 → Feb 26): $120.93 → $149.62 = +23.7% in one month.
SK Hynix crossed ₩1,000,000 on February 24 — closing at ₩1,005,000, up 5.68% in a single session. It was the fastest 1,000-point index jump in KOSPI history, taking just 34 days from 5,000 to 6,000.
Samsung at ₩214,500 is pacing toward a $1 trillion valuation. Investing.com's headline: "Samsung rally nears 200% as AI chip demand accelerates."
IV Update: The Mispricing Is Correcting — But Not Done
Current IV Snapshot
| Metric | Part 1 (Feb 23) | Today (Feb 26) | Change |
|---|---|---|---|
| EWY 30-Day IV | ~37% | 54.00% | +17 pts |
| IV Percentile | 84th | 98th | +14 pts |
| IV Rank | — | 91.03% | — |
| Historical Vol | 38.17% | 35.54% | -2.6 pts |
| IV High (recent) | — | 57.15% (Feb 17) | — |
| IV Low (52w) | — | 22.03% (Jun 2025) | — |
The 30-day IV has repriced from 37% to 54% — a 17-point jump. This is exactly the Vega expansion we described. But note: IV at 54% is still below our blended realized vol estimate of 60-65% from Part 1. The gap has narrowed but hasn't closed.
Critical observation: IV at 54% with historical vol at 35.54% means the options market is now pricing in FORWARD volatility above historical — a regime change from Part 1 where IV was below realized vol. The market is starting to wake up, but the 2028 LEAPS tenor still lags.
Put/Call volume ratio: 1.93 — heavy put buying suggests hedging demand, not directional bearishness. Institutional Korea longs are buying protection, which paradoxically supports call pricing by keeping skew steep.
Earnings Catalyst Confirmed: Record Numbers
SK Hynix FY2025
- Q4 Revenue: ₩32.8T (+34% QoQ, +66% YoY) — quarterly record
- Q4 Operating Profit: ₩19.2T (+68% QoQ, +137% YoY) — quarterly record
- Full Year OP: ₩47.2T — first time ever beating Samsung
- HBM revenue more than doubled YoY, HBM3E 12Hi driving volumes
- Locked down >2/3 of HBM supply orders for Nvidia's Vera Rubin
Samsung Electronics FY2025
- Q4 Revenue: ₩93.84T (+23.8% YoY)
- Q4 Operating Profit: ₩20.07T (+209% YoY) — quarterly record
- Memory division: ₩44T revenue, ₩16.4T OP in Q4 alone
- HBM4 finalized at 11.7Gbps — mass shipments to Nvidia started February 2026
- 1c DRAM yields above 80% — stable mass production confirmed
Combined
SK Hynix's full-year profit beat Samsung's for the first time in history (₩47.2T vs ₩43.6T). But Samsung's Q4 was actually larger than SK Hynix's Q4 — the HBM4 ramp is closing the gap.
2026 consensus: Citi projects Samsung's 2026 OP at ₩251T (~$173B) — more than Apple's 2025 operating profit alone.
Forward Catalysts Update
Confirmed Since Part 1
-
Samsung HBM4 shipments to Nvidia began February 2026 — no longer forward, it's happening now. Samsung allocating up to 150K wafers/month of 1c DRAM to HBM4 by year-end (+47% capacity).
-
DRAM pricing stronger than expected — ASPs jumped 30%+ QoQ in Q4, another 40-50% increase projected through H1 2026. HBM4 priced in mid-$500 range — double HBM3E.
-
KOSPI structural breakout — 5,000→6,000 in 34 days, fastest ever. Rally broadening beyond semis into financials and autos.
-
HBM3E price hikes — Samsung and SK Hynix both implementing ~20% increases for 2026.
Still Ahead
- Nvidia Vera Rubin launch (H2 2026) — next-gen AI accelerator requiring HBM4
- 2x leveraged Samsung/SK Hynix ETFs (June 2026) — structural vol injection ~4 months out
- Samsung $1T valuation — psychological milestone triggering index reweighting
EWY Call Trade: Updated Positioning
What's Working
IV expanded from 37% to 54% → Vega gain. Underlying up 7.4% → Delta gain. Jan 2028 $160 calls (14.9% OTM at Part 1) are now only 6.9% OTM.
Updated Risk Assessment
Higher risk now:
- IV at 98th percentile — near-term Vega tailwind mostly captured
- EWY up 158% in 12 months — mean reversion risk is real
- Won strength could reverse on global risk-off
Still favorable for 2028 LEAPS:
- 2028 tenor IV lags the 30-day IV reprice — term structure inverted, long-dated vol cheaper
- $173B Samsung OP estimate for 2026 not fully priced at $149
- HBM4 ramp through 2027 — real revenue acceleration still ahead
- Vera Rubin + 2x ETFs create future vol catalysts
Position Management
Existing positions (entered near Part 1 at ~$139):
- Take partial profits on near-dated calls (2026 expiry) — IV expansion captured
- Hold 2028 LEAPS — forward catalysts haven't fired yet
- Consider rolling strikes up to $180-$200 to maintain OTM convexity
New entries at $149:
- Smaller sizing: 0.5-1% of portfolio (down from 1-3% in Part 1)
- Prefer Jan 2028 $180-$200 strikes (20-35% OTM)
- The trade is now more directional than vol-arb at current IV levels
Samsung vs SK Hynix: The Convergence Trade
SK Hynix dominated HBM in 2025 (57% revenue share vs Samsung's 22%). But Samsung's Q4 memory OP actually exceeded SK Hynix's, and HBM4 shipments have begun.
If Samsung closes the HBM gap — 150K wafers/month of HBM4 capacity suggests they will — the 28% Samsung weighting in EWY gets a double boost: memory supercycle AND market share recovery. SK Hynix leading was good; Samsung catching up is better for the ETF.
TLDR
- EWY $139.30 → $149.62 since Part 1 (+7.4%), +24% month-to-date
- IV repriced 37% → 54% (98th percentile) — Vega captured, but 2028 LEAPS term structure still cheap
- SK Hynix ₩1,000,000 — first Korean stock at seven figures, ₩47.2T annual profit
- Samsung ₩214,500 — pacing toward $1T, HBM4 shipping to Nvidia NOW
- KOSPI 6,000 — fastest millennium-point jump in history
- Existing holders: hold 2028 LEAPS, take partial on near-dated, roll strikes up
- New entries: smaller size, higher strikes ($180-$200), directional bias over vol-arb
- Key catalysts ahead: Vera Rubin H2 2026, 2x leveraged ETFs June 2026, Samsung HBM4 ramp
第二部分更新 — 2026年2月26日
原始报告发布于2026年2月23日。EWY收盘价$139.30。今日：$149.62（+7.4%）。
原始报告发布以来，KOSPI首次突破6,000点历史大关。SK海力士股价突破₩1,000,000——韩国史上首只达到七位数的股票。三星于2月24日突破₩200,000，今日交易价₩214,500。EWY本月累计上涨24%，自Part 1以来上涨7.4%。
波动率套利论点正在兑现。以下是最新数据。
Part 1以来的价格走势
| 标的 | Part 1（2月23日） | 今日（2月26日） | 涨幅 |
|---|---|---|---|
| EWY | $139.30 | $149.62 | +7.4% |
| 三星 (005930.KS) | ₩193,000 | ₩214,500 | +11.1% |
| SK海力士 (000660.KS) | ₩949,248 | ₩1,018,000 | +7.2% |
| KOSPI | ~5,400 | 6,000+ | +11.1% |
EWY月度走势（2月2日→2月26日）：$120.93 → $149.62 = 一个月+23.7%。
SK海力士于2月24日突破₩1,000,000大关——收盘₩1,005,000，单日涨幅5.68%。KOSPI从5,000到6,000仅用34天，创历史最快千点突破纪录。
三星₩214,500正迈向1万亿美元市值。Investing.com标题："三星涨幅接近200%，AI芯片需求加速。"
隐含波动率更新：错误定价正在修正——但尚未完成
当前IV数据
| 指标 | Part 1（2月23日） | 今日（2月26日） | 变化 |
|---|---|---|---|
| EWY 30日IV | ~37% | 54.00% | +17个百分点 |
| IV百分位 | 84th | 98th | +14 |
| IV Rank | — | 91.03% | — |
| 历史波动率 | 38.17% | 35.54% | -2.6 |
| IV近期高点 | — | 57.15%（2月17日） | — |
| IV 52周低点 | — | 22.03%（2025年6月） | — |
30日IV从37%重新定价至54%——跳升17个百分点。这正是我们描述的Vega扩张。但注意：54%的IV仍低于Part 1估算的60-65%混合实际波动率。差距缩小但未消除。
关键观察：IV 54%而历史波动率35.54%，意味着期权市场现在定价的是高于历史水平的前瞻性波动率——这是相对Part 1中IV低于实际波动率的体制性转变。市场正在觉醒，但2028 LEAPS长期限仍然滞后。
看跌/看涨成交量比：1.93——大量看跌期权买入表明对冲需求，而非方向性看空。
财报催化剂确认：创纪录数据
SK海力士 FY2025
- Q4营收：₩32.8T（+34% QoQ，+66% YoY）——季度纪录
- Q4营业利润：₩19.2T（+68% QoQ，+137% YoY）——季度纪录
- 全年营业利润：₩47.2T——历史首次超越三星
- HBM收入同比翻倍以上，锁定Nvidia Vera Rubin 超过2/3的HBM供应订单
三星电子 FY2025
- Q4营收：₩93.84T（+23.8% YoY）
- Q4营业利润：₩20.07T（+209% YoY）——季度纪录
- HBM4已定型（11.7Gbps），2026年2月开始向Nvidia量产出货
- 1c DRAM良率超过80%——稳定量产确认
2026共识：花旗预测三星2026年营业利润₩251T（~$1,730亿）——超过苹果2025年营业利润。
前瞻催化剂
Part 1以来已确认
- 三星HBM4已于2026年2月开始向Nvidia出货——不再是预期，已成现实
- DRAM价格强于预期——Q4 ASP环比跳涨30%+，H1 2026再涨40-50%。HBM4定价$500+，是HBM3E的两倍
- KOSPI结构性突破——涨势从半导体扩散至金融、汽车板块
- HBM3E涨价——三星和SK海力士均实施~20%价格上调
未来催化剂
- Nvidia Vera Rubin发布（2026年H2）——需要HBM4
- 三星/SK海力士2倍杠杆ETF（2026年6月）——结构性波动率注入
- 三星1万亿美元市值——触发指数再平衡和被动资金流入
EWY看涨期权交易：更新仓位建议
现有仓位（Part 1附近~$139入场）
- 对近期到期合约（2026年到期）部分获利了结——IV扩张已兑现
- 持有2028 LEAPS——前瞻催化剂尚未触发
- 考虑将行权价上移至$180-$200以保持OTM凸性
新建仓位（$149入场）
- 更小仓位：组合的0.5-1%（Part 1建议为1-3%）
- 首选2028年1月$180-$200行权价（20-35% OTM）
- 当前IV水平下，该交易更偏方向性而非波动率套利
三星 vs SK海力士：收敛交易
SK海力士主导2025年HBM市场（收入份额57% vs 三星22%）。但三星Q4存储器营业利润实际超过SK海力士，且HBM4出货已开始。
如果三星缩小HBM差距——15万片/月HBM4产能表明他们会——EWY中28%的三星权重将获得双重提振：存储器超级周期+市场份额回升。SK海力士领先是好事；三星追赶上来对ETF更好。
总结
- EWY自Part 1以来$139.30→$149.62（+7.4%），月度+24%
- IV从37%重新定价至54%（98百分位）——Vega已兑现，但2028 LEAPS期限结构仍然便宜
- SK海力士₩1,000,000——韩国首只七位数股票，年利润₩47.2T
- 三星₩214,500——迈向$1T市值，HBM4正在向Nvidia出货
- KOSPI 6,000——史上最快千点突破
- 现有持仓：持有2028 LEAPS，近期合约部分了结，上移行权价
- 新建仓位：更小仓位，更高行权价（$180-$200），方向性为主
- 关键催化剂：Vera Rubin（2026 H2）、2倍杠杆ETF（2026年6月）、三星HBM4产能爬坡