Report Date : Aug 21, 2026 | Analyst : Invest101 Copilot
Ticker: SKHY (Nasdaq ADR) vs. 000660.KS (Korea Exchange) | Current Price: $163 (as of Aug 20, 2026 close) | Market Cap: ～$1.19 Trillion | Risk Level: High
Next Earnings: ~Oct 27, 2026 (Q3 FY26)
This report is for informational purposes only and does not constitute investment advice. Cross-border and short-selling strategies involve significant risk, including unlimited loss potential on shorts and foreign-exchange exposure.
Executive Summary
SK hynix executed the largest ADR listing in history (~$26.5B raised) on July 10, 2026, listing 177.9 million American Depositary Receipts on the Nasdaq under SKHY. Each ADR represents one-tenth (0.1) of a Korean common share (10 ADRs = 1 share of 000660.KS). [Source: Reuters via Yahoo Finance, Bloomberg via Yahoo Finance]
The ADR currently trades at roughly a +30% premium to the identical economics of the Seoul-listed common shares. That premium has swung between +16% and +51% since listing and averaged ~+26%. [Source: WSJ via Investing.com, Bloomberg]
This premium is real, it is visible, and it is NOT risklessly arbitrageable today. SK hynix has capped the number of common shares that can be converted into ADRs at 2.5% of shares outstanding, and that quota is already fully utilized (per the Korea Securities Depository). The conversion direction you would need to monetize the premium — buy cheap Seoul shares → convert into ADRs → sell expensive ADRs — is effectively closed. The only open direction is ADR→common (cancellation), which is unprofitable when the ADR sits at a premium. [Source: Bloomberg via Yahoo Finance]
Practical implication: the “arbitrage” is really a relative-value pair trade (short SKHY / long 000660.KS) with no convergence guarantee. It can (and has) bled wider. For long investors, the Korean line is simply the ~30%-cheaper way to own identical economics.
The Cross-Listing Structure
SKHY is a sponsored, capital-raising ADR program (not an unsponsored Level 1), with Citigroup as depositary bank. The offering sold 177.9 million ADRs (= 17.79 million new common shares) at $149 per ADR, raising ~$26.5B — larger than Alibaba’s 2014 debut. The ADR debuted July 10, opened ~14% above the offer at $170, and closed its first day at $168. [Source: Reuters, Barron’s via Yahoo Finance]
Key structural facts that drive everything else:
| Parameter | Detail |
|---|---|
| ADR ratio | 1 ADR = 0.1 common share (10 ADRs = 1 share) |
| Depositary | Citigroup |
| ADR float | ~177.9M ADRs ≈ $29B (tiny vs. ~$860–890B company) |
| Conversion cap (common→ADR) | 2.5% of shares outstanding — FULLY UTILIZED |
| Conversion direction open | ADR→common (surrender/cancellation), ~3–5 days, fees |
| Underwriters | BofA, Citi, Goldman, J.P. Morgan (+9 others) |
| Use of proceeds | Yongin fab cluster (Y1) + ASML EUV lithography tools |
The small ADR float (~2.5% of shares, ~$29B) relative to the ~$860B+ total company is the single most important driver of the premium — this is a scarcity premium, not a fundamental one. [Source: Reuters, Bloomberg via Yahoo Finance]
Quantifying the Premium (Live)
Using live prices and FX (Aug 20–21, 2026):
| Input | Value | Source |
|---|---|---|
| 000660.KS price | KRW 1,732,000 | Yahoo Finance ticker |
| SKHY price | $163.08 | Yahoo Finance ticker |
| USD/KRW | 1,381.83 | Yahoo Finance FX |
| Implied ADR value = (1,732,000 ÷ 10) ÷ 1,381.83 | $125.34 | calc |
| ADR premium | +30.1% | calc |
Cross-check with an alternate intraday snapshot (000660 @ KRW 1,662,000, SKHY @ $156.48, FX ~1,390) gives +30.9% — so the ~30% premium is robust, not a timestamp artifact. [Source: Yahoo Finance]
Historical premium trajectory: ~+16% at first-day close, +25.6% after the July 13 Seoul rout, ~+29% at a Friday close in mid-July, a peak of ~+51% three days after listing, and ~+22% by late July. The current ~30% is toward the high end of the range. [Source: Reuters, WSJ via Investing.com, Bloomberg via Yahoo Finance]
Why the Premium Exists
The premium is the sum of a 「small justifiable component」 plus a 「large speculative/scarcity component」:
Justifiable (~a few points, at most ~15% by TSMC’s precedent):
- Dollar denomination — no FX conversion friction for USD investors (though the ADR still moves with won).
- No Korean securities transaction tax on US-side trades (Korea levies a tax on local sales).
- US settlement/custody convenience (T+1, standard brokerage).
- US index/ETF demand and tax treatment for US funds.
Excess (the other ~15–20 points):
- Scarcity: only ~2.5% of the company exists in ADR form, and that pool cannot be expanded because the cap is full.
- Speculative AI demand for a “pure-play” US-listed HBM leader — US retail/institutions wanting Nvidia-adjacent exposure without a Korean account.
- One-way conversion (ADR→common only) removes the arbitrage mechanism that would normally pin the two lines together.
The Bloomberg quote from Acadian’s Owen Lamont captures the mechanism precisely: “If you had two-way conversion, then you can’t have a premium… But once conversion is stopped in one direction, that opens up the possibility of a mispricing, and there’s no guarantee how long it would last.” [Source: Bloomberg via Yahoo Finance]
Benchmark: TSMC’s ADR averaged only a 3.2% premium over 2010–2020, rising to ~15–17% post-ChatGPT. SKHY at ~30% is roughly 2x TSMC’s elevated AI-era premium — a sign of how stretched the scarcity premium is. [Source: Bloomberg via Investing.com]
Verdict: No Riskless Arb, Only a Blocked Relative-Value Trade
Here is the mechanics, step by step, because this is the crux of your question.
Capture the premium ❌
Buy 000660.KS (cheap) → deposit shares with the custodian → have Citi issue new ADRs → sell SKHY (expensive) → pocket ~30%.
Why this doesn’t work at this moment: the common→ADR issuance channel is capped at 2.5% of shares outstanding and the quota is already 100% used. No new ADRs can be created until existing ADR holders cancel receipts and free up quota. So the “convert and deliver” leg that would lock the trade is unavailable. [Source: Bloomberg via Yahoo Finance]
The reverse trade (to capture a discount) ❌
Buy SKHY (cheap) → surrender to Citi → receive Korean shares → sell 000660.KS (expensive). This direction is open (cancellation is uncapped). But it is only profitable if the ADR trades at a discount — and it trades at a premium. So this channel is dormant.
What’s left — a “statistical” pair trade ✅
Short SKHY + long 000660.KS, betting the premium to narrow. This is not riskless:
- No convergence guarantee — with the conversion channel blocked, the premium can widen (it went from 16% to 51% in three days). A short-ADR position is exposed to further premium expansion.
- Borrow scarcity/cost — SKHY is newly listed with a tiny float; borrow can be expensive or unavailable, and squeezes are a live risk.
- FX and timing mismatch — Seoul trades in a different time zone; you carry won risk unless hedged.
- Korean-side frictions — foreign-investor registration, local transaction tax on the Seoul leg, dividend withholding.
Conclusion: There is no mechanical arbitrage. What exists is a directional bet that the premium compresses. That bet has historically been a coin-flip in the near term (the premium is volatile) but has a strong structural tailwind over a 6–12 month horizon (see §5). This is an institutional relative-value trade, not a retail arbitrage. [Source: Bloomberg via Yahoo Finance, WSJ via Investing.com]
What Breaks the Premium (Compression Catalysts)
The premium’s persistence depends entirely on the 2.5% cap staying binding. Watch these:
- Cap increase or fresh ADR issuance. SK Group Chairman Chey Tae-won said on July 10 the company is “open to issuing more ADRs if investor returns are strong and the share price remains stable.” Any increase in the ADR float would open the common→ADR channel and mechanically compress the premium. This is the single most important catalyst. [Source: Bloomberg via Yahoo Finance]
- ADR holder cancellations (ADR→common) freeing up quota — possible if holders want the underlying shares for the buyback/cancellation program or to arbitrage the other way.
- Korean share outperformance — the premium narrows mechanically if Seoul rallies while the ADR lags (already partly the case: 000660 +156% YTD vs. the ADR’s short, choppy history).
- Weaker US AI-sentiment — a risk-off in US AI names hits the speculative ADR premium hardest.
Base case: expect the premium to mean-revert from ~30% toward the 15–20% TSMC-style band over 6–12 months, most likely via ADR underperformance rather than Seoul outperformance. A 20% premium implies an ADR price of ~$150; a 15% premium implies ~$144; full convergence to fair value (~5% premium) implies ~$132 — all versus the current $163.08. That is the risk embedded in holding the ADR rather than the Korean line. [Source: calc, Bloomberg]
👨🚀 Top AI-Powered Smart Stock Analyst
Have no clue on how to analysis a stock or ETF? Let Invest101 Copilot do the heavy lifting for you !
Simply enter a stock ticker, ETF, or company name, and it will instantly deliver an institutional-grade research report in 2 minutes. Get the in-depth data and actionable insights you need to make smarter investment decisions, faster. Learn more
Investment Thesis — Bull vs. Bear
Bull case: SK hynix is the monopoly-quality franchise at the heart of the AI buildout — ~58% HBM share, sold-out order book, 76%+ operating margins, net cash, and now returning >50% of FCF via buybacks. At 4.5x forward earnings (vs. Micron ~6.7x), the Korean line is cheap even on materially discounted mid-cycle earnings. The ADR premium, while rich, reflects genuine scarcity and index/ETF demand that can persist (TSMC proved 15–17% can last for years). [Source: Reuters, Bloomberg, Yahoo Finance]
Bear case: these are peak-cycle earnings; 2027–28 supply additions risk a memory downturn, and the “circular AI financing” narrative questions demand durability. On the ADR specifically, ~30% is a double premium to TSMC’s already-elevated level, and the very mechanism sustaining it (the 2.5% cap) is one board decision away from being lifted — which would trigger a sharp, possibly sudden premium collapse. Holding SKHY instead of 000660.KS is, in effect, paying ~30% extra for the same cash flows and taking disproportionate downside. [Source: Bloomberg, WSJ via Investing.com]
Net: constructive on the underlying franchise at these multiples; structurally negative on the ADR’s premium relative to the Seoul line.
Business Model & Financial Health
SK hynix is the world’s #1 high-bandwidth memory (HBM) supplier with ~57–58% HBM revenue share (Counterpoint, Q1 2026), supplying the majority of the HBM used in Nvidia’s AI accelerators. Its mix spans DRAM (the profit engine, HBM within it), NAND flash/SSD, and MCP/foundry. Growth is being driven by HBM3E/HBM4 and AI-server demand, with its HBM order book effectively sold out through 2026. [Source: Reuters, Bloomberg via Yahoo Finance]
Financial snapshot (underlying company, KRW — from 000660.KS key statistics, Aug 14, 2026):
| Metric | Value |
|---|---|
| Revenue (TTM) | KRW 189.17T (~$137B) |
| Net income (TTM) | KRW 162.08T (~$117B) |
| Gross margin | ~70% |
| Q1 FY26 revenue / net earnings | revenue KRW 52.58T (annualized +198% YoY), net income KRW 40.33T. [Source: Yahoo Finance, SEC F-1 via Yahoo] |
| Q2 FY26 revenue / op. margin | revenue KRW 79.3T (+257% YoY), ~76% operating margin on HBM3E/AI demand. [Source: Polymarket event metadata] |
| Dividend | interim KRW 375/share, ex-date Aug 28, 2026; forward dividend KRW 1,500/yr (~0.09% yield). [Source: Yahoo Finance] |
| Trailing P/E / Forward P/E | 15.6x / 4.5x |
| ROE / ROA (TTM) | 92.7% / 33.7% |
| Total cash / total debt | KRW 87.96T / 18.59T (net cash ~$50B) |
| Debt/Equity | 7.1% |
| Current ratio | 17.5x |
| Next earnings report date | ~Oct 27, 2026 (Q3 FY26, per Yahoo Finance earnings calendar estimate) |
Data caveat: Yahoo’s headline “net margin” of 85.7% and operating margin (76.3%) exceeding gross margin (70.2%) are internally inconsistent — a known artifact of mixing TTM and peak-quarter figures in the ADR feed. Treat the Q1 FY26 net margin (~77%) and Q2 op. margin (~76%) as the cleaner reference points. [Source: Yahoo Finance]
Balance sheet health: exceptional. Net cash of ~$50B, negligible leverage (D/E 7%), extreme liquidity (current ratio 17.5x), and a ~$26.5B equity raise layered on top. The operating cash flow (TTM) is ~KRW 127.2T (~$92B) — the company is decisively, sustainably cash-flow positive. The new KRW 40T (~$29B) treasury buyback-and-cancel program (announced Aug 19, 2026) commits over half of projected 2025–27 free cash flow to shareholders — a major capital-return pivot. [Source: Yahoo Finance, Polymarket event metadata, Reuters]
Risk Factors
- ADR premium compression — the #1 idiosyncratic risk to SKHY specifically (detailed above). Holding the ADR embeds ~$30–38/share of pure premium that can evaporate.
- Memory cyclicality / 2027–28 oversupply — Morningstar flags that new capacity coming online in 2027–28 could flip today’s shortage into price erosion. CEO Kwak counters that 2027 will see the most severe shortage in memory history — a genuine bull/bear debate. [Source: Reuters]
- “Circular AI financing” fears — Nvidia’s ~$750B in deals and customer equity/debt funding have reignited sustainability questions about AI capex; this triggered the July 28 global chip rout. [Source: Bloomberg via Yahoo]
- China competition & geopolitics — export controls, Chinese memory entrants, and US-Korea policy risk.
- HBM4 ramp execution — NH Investment flagged HBM4 shipment scale-up lagging expectations in Q2. [Source: Reuters]
- FX — a strengthening won mechanically reduces the USD value of the underlying (and thus the ADR), while also narrowing the premium.
Catalysts & Sentiment
- Record KRW 40T buyback/cancellation (announced Aug 19, 2026) — already drove a +4% pop on Aug 20 while the broader tech complex fell. Execution is a near-term tailwind. [Source: 24/7 Wall St., Polymarket]
- Q3 2026 earnings (~Oct 27) — HBM4 ramp, conventional DRAM pricing, and FY27 supply commentary.
- Possible ADR cap increase / follow-on — the key premium-compression catalyst (§5).
- Samsung’s ~$70B+ shareholder-return package (board meeting Aug 21) — memory-sector capital-return momentum is a rising tide for the group. [Source: Reuters]
- Yongin fab and Japan Miyagi plant news — capacity expansion narrative. [Source: Reuters]
- Sentiment: consensus is Strong Buy (13 analysts on the ADR, mean rating 1.43; 40 on the Korean line, 1.33). Korean-line target ~KRW 3.17M (implied ~83% upside); ADR target ~$245 (implied ~50% upside, which itself assumes the premium persists). The debate is entirely about peak-earnings sustainability vs. HBM structural demand, with the ADR premium as a secondary overlay. [Source: Yahoo Finance]
Competitive Landscape & Related Stocks
| Company | Ticker | HBM share (Q1 26) | Note |
|---|---|---|---|
| SK hynix | 000660.KS / SKHY | ~58% | #1 in HBM |
| Samsung Electronics | 005930.KS | ~21% | #2 memory player |
| Micron | MU | ~21% | US HBM pure-play |
| Kioxia | 285A.T | — | NAND competitor |
| SanDisk / Western Digital | SNDK / WDC | — | NAND exposure |
Technical Analysis & Entry
SKHY has only ~6 weeks of trading history, so classic indicators are immature (50-day and 200-day MAs both sit at ~$154.7). The pattern: IPO $149 → first-day $168 → all-time high $194.80 (Jul 14) → flush to $124.80 (Jul 29, the “circular AI” rout) → recovery to $163.08 (Aug 20, +4.4% on the buyback). [Source: Yahoo Finance, get_stock_prices]
Momentum: improving off the July low, price back above the ~$154.7 MA, but the tape is extraordinarily volatile (beta 2.4; multiple ±8–15% days). Resistance sits at the $177–180 area (recent Polymarket-struck highs) and then $194.80; support at $150–155 (MA + round number), then $140, then $124.80.
3-month entry suggestion: Do not chase SKHY here at a ~30% premium. For ADR longs, an optimal entry is $140–150 (which would still embed a ~12–20% premium), or $125–135 on a risk-off flush toward the July low. Better yet, if your broker offers KRX access, buy 000660.KS instead — identical economics at a ~30% discount, with the added kicker that the Korean line has more upside to the ~KRW 3.17M consensus target. Use SKHY weakness caused by premium compression as the signal to reassess. [Source: Yahoo Finance, calc]
Valuation
A rigorous DCF here is limited by (a) only ~2 quarters of clean post-listing fundamental data and (b) peak-cycle earnings that would flatter any terminal-value math. I anchor instead on a normalized-earnings cross-check:
- Even assuming 2027 memory earnings normalize down ~60% from the current run-rate (~KRW 190T annualized net income → ~KRW 75T), the Korean line at ~KRW 1.73M would trade at roughly 10–11x normalized earnings — reasonable-to-cheap for a franchise-quality #1 HBM supplier with net cash.
- Forward P/E of 4.5x on the Korean line vs. 6.66x for Micron means SK hynix trades at a ~33% discount to Micron despite ~2.7x Micron’s HBM share — the market is pricing peak-earnings decay aggressively. [Source: Yahoo Finance, Reuters]
- The ADR at ~30% premium is the only component that looks outright expensive. Consensus ADR target $245 effectively assumes the premium persists; a DCF/peer-multiple view of the underlying (KRW 2.5M–3.2M/share) supports ~$180–230 ADR fair value if a 15–20% structural premium is granted, but only ~$135–145 at a normalized 5–10% premium. Valuation verdict: the Korean line is attractive; the ADR’s premium is the overvaluation. [Source: calc, Yahoo Finance]
Options Strategies
SKHY has listed Nasdaq options (weekly chain through Aug 21 visible in the data; longer-dated series exist but the feed returned the near-dated weekly). Liquidity is thin and lumpy outside a few strikes, and implied vol is elevated (newly listed, high realized vol). Three idea structures, all framed around the premium-compression thesis:
- Bear put spread to fade the premium
- buy the $160 put + sell $140 put (or buy $150 / sell $130 put) expiring over the next 1–3 months.
- Rationale: the ~30% premium argues for mean-reversion toward $140–150. Defined risk, benefits from a compression-driven pullback without requiring a short stock borrow.
- Call credit spread / covered call on the upper range
- sell the $185–$190 call spread (or sell the $185 call against a long).
- Rationale: collect the high IV while the stock is pinned below $180 resistance; profits if SKHY stays range-bound or the premium only slowly decays. This is the “earn carry while the premium persists” expression.
- Collar / risk-reversal for existing ADR holders
- if already long SKHY, sell an OTM call (~$185) and buy an OTM put (~$140) to fund downside protection against the one catalyst (cap lift / follow-on issuance) that could gap the premium lower overnight.
Caveat: all SKHY option liquidity is shallow — use limit orders and modest size. [Source: Alpaca options chain]
Polymarket Signals
Polymarket hosts a weekly “SKHY Hit Price” series (resolves via Pyth 1-minute candles). Signals are thin but directionally informative:
- For the week of Aug 17: the $177, $174, $171, $168 HIGH strikes all resolved YES (the stock traded through them), and the $165/$162/$159/$156 LOW strikes also resolved YES — confirming the wide, choppy range.
- The $180 HIGH strike was pricing only ~7% probability and $186 ~50% (illiquid/spread), indicating the market sees $180+ as unlikely in the near term.
- Most usefully, the market’s own event metadata frames the drivers: Q2 revenue +257% YoY to KRW 79.3T with ~76% op margin, plus the KRW 40T buyback as the sentiment catalyst.
Net signal: no tradable edge beyond confirming the current $150–$180 chop, with the market fading the idea of an imminent breakout above $180. [Source: Polymarket]
Institutional Ownership
Because SKHY listed July 10, 2026 (Q3), it will not appear in 13F filings until the Q3 2026 reporting cycle (due November 2026) — so current 13F data is essentially empty. What we know from the deal itself: three anchor investors — Baillie Gifford, Coatue Management, and Situational Awareness Partners — signaled up to ~$7B of combined appetite, and the book was >7x oversubscribed with long-only, tech, sovereign-wealth, and Asia-focused funds participating. [Source: Reuters via Yahoo Finance]
Bottom Line
SKHY vs. 000660.KS is not a mispricing you can harvest — it is a structurally engineered premium. The ~30% gap exists because the arbitrage channel (common→ADR creation) is capped at 2.5% of shares and already full, so the normal force that would pin the two lines together is switched off. There is no riskless conversion arbitrage; only a directional relative-value trade (short SKHY / long 000660.KS) whose payoff depends on premium compression and whose risk is further premium expansion (it has already hit +51%).
The asymmetry is the opportunity: the premium is roughly double the elevated TSMC-era norm (15–17%), is supported by a cap that SK Group’s own chairman has said he is “open” to raising, and would likely normalize to ~$140–150 ADR (a 12–20% premium) if that cap lifts or fresh ADRs are issued. For longs, buy the Korean line, not the ADR — you get identical economics ~30% cheaper and sidestep the single biggest idiosyncratic risk in the ADR. The underlying franchise itself — #1 HBM share (~58%), ~76% operating margins, net cash, a record buyback, and a 4.5x forward multiple — remains compelling at the Seoul price. The premium, not the company, is what’s stretched.
| Disclaimer: Content provided by Invest101 Copilot is AI-generated and may contain outdated, inaccurate, or misleading information. This tool is for research assistance only and does not constitute investment advice. Invest101 is not responsible for any consequences arising from the use of this content. Investors should conduct their own due diligence. |
