$SKHYNASDAQSK hynix Inc.initiation · Q2 2026 · as of 2026-08-06 · reported 2026-07-29

SK Hynix sold Wall Street the biggest foreign listing in US history, printed a record quarter three weeks later, and the ADR trades two dollars above the offer price.

Priced at $149, public at $168, $194 within three sessions, $127 after a record quarter that somehow disappointed, and $151 today: four weeks of SKHY compressed a whole memory cycle into one tape. Underneath the noise sits the company that supplies roughly 57% of the world's HBM, earning a 76% operating margin, at about seven times trailing earnings.

Market Cap
$1.07T
10 ADRs equal one Seoul-listed share
Vs Offer Price
+1.4%
177.9M ADRs priced at $149, a $26.5B raise
Q2 Revenue
+257% YoY
₩79.3T, about $55.2B
Trailing P/E
7.3x
under 4x next year's street EPS
01 — The Print

Q2 2026 earnings, decoded

The first print as a US-listed name was a record on every operating line and still landed about ₩5 trillion short of an AI-charged consensus. Dollar figures use the roughly ₩1,437 per dollar rate carried in wire coverage of the release.

Revenue
₩79.3T
+257% YoY
$55.2B; up 51% QoQ from ₩52.6T
Operating Profit
₩60.5T
76% margin
$42.2B, up 557% YoY; both all-time records
Gross Margin
83%
+4 pts QoQ
DRAM ASPs up ~30% QoQ, NAND up mid-50s
Net Income
₩93.9T
118% of revenue
$65.4B; includes ₩63.3T of one-off gains on the Kioxia stake
The Miss
-₩4.7T
-6% vs consensus
the street wanted ~₩84T revenue and ~₩64T operating profit
Follow every dollar — the income statement, as a flow
Ribbon width scales with dollars. Slate = segments, green = profit, red = costs.
DRAM (73%) $40.3BNAND (27%) $14.9BRevenue $55.2BGross Profit $45.8BCost of Revenue $9.4BOperating Profit $42.1BOperating Expenses $3.7B
Line by line — with the why
Bars scaled to revenue. Click a row for the read.
₩79.3T, the first quarter fully on AI contract pricing; H1 revenue crossed ₩100T for the first time
Only 17 cents of every revenue dollar (implied from the deck's 83% gross margin); ASPs repriced far faster than fab costs
An 83% gross margin, up 29 points in a year, per the Q2 results deck
R&D-heavy opex (implied from gross and operating profit) is now a rounding error against contract-priced revenue
₩60.5T at a 76% margin; the chart stops here because one-off Kioxia gains push net income above revenue
Guidance
Q3 DRAM Bit Shipments+~10% QoQ
Q3 NAND Bit Shipmentslow single digit QoQ
HBM4full-scale ramp in H2 2026mass shipments began in Q2
FY2026 Capexhigh-₩40T range
2026 DRAM Industry Demandmid-20s% bit growth
Balance sheet
Cash & Equivalents₩88.0T (~$61B)
Total Debt₩18.6T (~$13B)
Net Cash₩69.4T (~$48B)
Debt-to-Equity7%
02 — The Ecosystem

Who SK buys from & sells to

More than half the world's HBM flows through this map: TSMC and the toolmakers feed the fabs, Nvidia and the hyperscalers drain them. Click any node.

The value chain, mapped
Hover or tap a node to light up its links
ASMLASML HoldingTSMTSMCAMATApplied Materials042700.KSHanmi SemiconductorNVDANvidiaMSFTMicrosoftAMDAMDGOOGLAlphabet$SKHYthe pivot
Tap a ticker — and I'll tell you exactly why it rises (or doesn't) with $SKHY.

Flow weight is a subjective 1-10 read of how tightly each name's results move with SK hynix's memory cycle, not a revenue rank.

↑ SKHY BUYS FROM
SKHY SELLS TO ↓
The memory complex
MU Micron Technology005930.KS Samsung ElectronicsSNDK SanDisk285A.T Kioxia Holdings
The HBM toolchain
LRCX Lam ResearchKLAC KLABESI.AS BE Semiconductor
03 — The Debate

Bull, base & bear case

Twelve initiations agree this is the best memory asset on earth; the tape's first month says owning it will not feel that way. The argument is price and cycle, not the franchise. Tap a case.

“You can now buy more than half the world's HBM supply for less than four times next year's earnings.”
The crown asset, priced like a commodity
  • 57% of the HBM market Roughly 57% share against Micron's ~21%, with HBM4 already shipping into Nvidia's Vera Rubin platform at yields nearing HBM3E maturity.
  • Ten five-year contracts About 10 long-term agreements of roughly five years lock in the order book, and industry voices see the memory bottleneck persisting into 2028.
  • The miss was a delay, not demand Q2's shortfall traced to HBM4 shipments sliding into H2 and conservative pricing; that revenue arrives this half as the ramp scales.
  • A fortress after the raise ₩69.4T of net cash and a 7% debt-to-equity ratio: the $26.5B ADR proceeds fund the capex war without touching leverage.
BULL CASE · 12-MO TARGET
$300
≈ +99% vs $151
Assumes the HBM4 ramp lands on schedule, next-year EPS near the ~$42 per ADR consistent with the street's sub-4x forward multiple, and a still-modest ~7x on it. Lands exactly on Barclays' post-cut $300, with Rosenblatt's $320 just above.
SUBJECTIVE ODDS
~30%
The three targets, side by side
12-month targets vs the $151.03 close. The spread tells you how two-sided this is.
BULL
$300
BASE
$200
BEAR
$115
Today ≈ $151.03
04 — Build Your Own

Price-target calculator

Pick a next-twelve-months EPS per ADR and the multiple you would pay for it. The whole SKHY debate fits inside these two sliders.

Implied value = EPS × P/E multiple

IMPLIED SHARE PRICE
$147
≈ -3% vs $151

Price target = forward EPS × P/E. The defaults ($21 of EPS at 7x) roughly reproduce today's $151 ADR price, matching the trailing $20.64 EPS and 7.3x multiple reported by StockAnalysis. The street's sub-4x forward multiple is consistent with next-year EPS above roughly $38, which is where the bull case lives; the bear case assumes earnings normalize toward the mid-teens as the 2027 supply wave lands.

05 — The Street

What Wall Street says

Coverage of the ADR is four weeks old and already emphatic: twelve analysts in the S&P Global poll, eleven of them at buy or better, with not a single sell.

Strong Buy
12 analysts · as of 2026-08-04
92%BUY OR BETTER
Strong Buy
7
Buy
4
Hold
1
Sell
0
Strong Sell
0
HIGH TARGET
Street high, S&P Global poll via StockAnalysis · 2026-08-04
$355
AVERAGE TARGET
S&P Global consensus via StockAnalysis · 2026-08-04
$244.17
LOW TARGET
Street low, S&P Global poll via StockAnalysis · 2026-08-04
$152
Every number here was set since the July 10 listing: Barclays opened at $330 and trimmed to $300 on softer commodity DRAM assumptions, Rosenblatt sits at $320 and UBS at $204, and the $152 street low is one dollar above the price, so even the most cautious initiation could not bring itself to model downside. The caveat cuts both ways: a four-week-old consensus has never been through a memory downcycle.
06 — The Call

Final assessment

The first month of trading handed you the offer price back with a record quarter attached.

BUY · INITIATE

Start the position at the round trip to $149; add if the HBM4 ramp shows up in the October print.

CONVICTIONMedium-High
12-mo Base Target
$200
+32% from $151.03
Trailing P/E
7.3x
on $20.64 ttm EPS per ADR
Net Cash
₩69.4T
about $48B after the ADR raise
Odds-weighted Value
$209
across bull, base and bear

The listing gave US investors direct access to the memory cycle's crown asset: roughly 57% of the world's HBM, a 76% operating margin, about 10 long-term agreements of roughly five years, and ₩69.4T of net cash. The Q2 "miss" was a timing story, not a demand story: shipments of some high-value-add products slid into the second half, in management's framing, while consensus had them in the quarter, and the tape converted a record print into a 35% drawdown from the debut-week high. At about seven times trailing earnings, that is expectations repricing, not evidence against the franchise.

The honest caveats: this is the same memory cycle we underwrite in the Micron brief, and the exam is the same 2027-2028 supply wave, with SK hynix itself funding a capacity doubling while Korea's market structure keeps exporting flash-crash volatility into the ADR. Buy the round trip to the offer price, size the position for the cycle rather than the order book, and let the October print arbitrate the HBM4 timing question.

▲ WHAT PUSHES US MORE BULLISH
  • HBM4 lands in the Q3 print The pushed-out shipments arriving on schedule would flip the miss narrative and re-anchor consensus revenue higher.
  • Visibility extends into 2028 New long-term agreements or customer prepays that push booked capacity into 2028 would break the supply-wave bear case, for SKHY and Micron alike.
  • The cash mountain gets a policy Shareholders are already demanding bigger payouts; a formal return program on ₩88T of cash would force a re-rate on its own.
▼ WHAT FLIPS US CAUTIOUS
  • Hyperscaler capex flinches Any guided slowdown in AI infrastructure spend hits the ~57% HBM leader first and hardest.
  • Samsung gets aggressive in HBM4 Qualification wins or price cuts by the runner-up would test the conservative-pricing playbook that already cost SK hynix its Q2 consensus.
  • The structure keeps misfiring More Nextrade flash crashes or leveraged-product unwinds in Seoul would keep the ADR's risk premium wide no matter what the fabs earn.
  • Supply lands early Accelerated 2027 capacity from any of the big three plus CXMT would put the young contract framework to its first real stress test.
This brief is one analyst's opinion, produced for research and education. It is not financial advice and no figure here is a recommendation to buy or sell any security. All figures are as of 2026-08-06 unless dated otherwise, and every claim traces to the sources listed on this page. Do your own work.
Sources · 36