Save the PDF to draw one Owl Card (per stock, once a day)
OWL LABS EQUITY RESEARCH
000660 · SIGNATURE REPORT OWL-260729-000660 · JULY 29, 2026
SIGNATURE QUANT PROFILE · 반도체
SK hynix Inc. 000660 · KOSPI
반도체 · KRW listed
OWL VERDICT
BALANCED CASE
Grade B · Score 70 / 100
LAST PRICE
₩1,399,000
▼ -9.7%
OWL FAIR VALUE
₩2,453,243
+75.4% vs price
STREET TARGET
₩2,712,489
+93.9% · 38 analysts
MARKET CAP
₩993.1T
div yield 0.14%
P/E · P/B
31.6 · 11.3
ROE 35.6%
52-WEEK RANGE
245,000–2,987,000
now at 42th pctile
Investment Summary
01
Quality vs. price, measured. Quality axes (ROE 35.6%, op. margin 48.6%) score 40/40; value axes (P/E 31.6x, P/B 11.26x, EV/EBITDA, dividend) score 33/80. The composite 70/100 (grade B) is that arithmetic.
02
No tracked 13F position. None of the 9 tracked managers reports 000660 this quarter (13F covers US-listed instruments only) — see p.3 for each book's style lens.
03
2 of 5 value strategies pass. Greenblatt Magic Formula, Prasad Deep Value admit 000660; the other 3 reject on their coded rules (details p.2).
04
What would change this report. Model fair value is ₩2,453,243 (+75.4% vs price); The verdict changes only when the data changes — screening reruns every weekday.
THE OWL VIEW
The data reads sound but not one-sided: composite 70/100 — 2/5 value screens pass. We grade 000660 B; the framework says the case is solid and price is the swing variable.
This report aggregates Owl Labs quant modules — 7-axis screener, 5 value strategies, 9 guru 13F filings, fair-value model, technical signals and news. All figures are measured from data sources as of the date above; sentences are rule-based templates (no generative claims). The score is a fitness test of reported fundamentals against today’s price — a screener, not a forecast; forward growth is not part of the score. Not investment advice.
000660 · SIGNATURE REPORT OWL-260729-000660 · JULY 29, 2026
Valuation — Five Value Disciplines
SECTION 02 OF 5
Each strategy is a mechanical screen coded from the discipline's published rules — a stock either passes or it does not. 000660 passes 2 of 5. Verdicts are computed, not opined; the measured inputs are shown for audit.
STRATEGY
MECHANICAL CRITERIA
VERDICT
MEASURED INPUTS
Graham Defensive
Low P/E × low P/B composite screen
FAIL
P/E×P/B 356
Greenblatt Magic Formula
Earnings yield + return on capital, top 30%
PASS
E/P 3.2% · ROE 35.6%
Prasad Deep Value
Low P/B, low leverage, positive earnings
PASS
P/B 11.26x
Simple DCF
OWL LABS EQUITY RESEARCH
000660 · SIGNATURE REPORT OWL-260729-000660 · JULY 29, 2026
Smart-Money Positioning — 9 Tracked Managers
SECTION 03 OF 5
Holdings from SEC 13F filings (disclosed ~45 days after quarter end). Weight = % of the manager's disclosed US long book. 13F covers US-listed instruments only — a Korean-listed share is "not reportable", not "passed over". These are filed positions — not the managers' opinions on this stock.
MANAGER
FUND
STYLE
WEIGHT
ACTION
STYLE LENS (RULE-BASED)
Warren Buffett
Berkshire Hathaway Inc.
Value
—
NO POSITION
Out of 13F scope — US-listed instruments only
Ray Dalio
Bridgewater Associates, LP
Macro
—
NO POSITION
Out of 13F scope — US-listed instruments only
OWL LABS EQUITY RESEARCH
000660 · SIGNATURE REPORT OWL-260729-000660 · JULY 29, 2026
Tape, Catalysts & Risk Register
SECTION 04 OF 5
Technical Readout as of Tue Jul 28
RSI (14d)36.1 — neutral
MACD vs signal-49357.63 below
MA20 / MA60₩2,048,200 / ₩2,100,133
Cross stateNone
Distance to 52-wk high-53.2%
52-wk range position42th percentile
Day change-9.7%
OWL LABS EQUITY RESEARCH
000660 · SIGNATURE REPORT OWL-260729-000660 · JULY 29, 2026
Financial Record & Final Judgment
SECTION 05 OF 5
FISCAL YEAR
FY2023
FY2024
FY2025
Revenue
₩32.77T
₩66.19T
₩97.15T
Operating income
-₩7.73T
₩23.47T
₩47.21T
Net income
-₩9.11T
₩19.79T
₩42.92T
Op. margin
-23.6%
35.5%
48.6%
Net margin
-27.8%
29.9%
44.2%
ROE
-17.0%
26.8%
35.6%
20/20
Op. margin
20/20
EV/EBIT
10/20
Dividend
3/20
Attention
1/20
Composite (incl. adj.)70 / 100 · B
KR universe rank#159 of 1036 · top 15%
Value screens passed2 / 5
52-wk range position42%
Next earningsTBA
Axis weights are proprietary; axis names and raw scores are disclosed. Attention measures news & search interest — a crowding proxy, not a quality signal.
One-line read: 반도체 name scoring 70/100 (grade B), passing 2/5 value strategies.
Conservative discounted cash flow vs price
FAIL
Model ₩2,453,243 vs ₩1,399,000
Damodaran Relative
Multiples vs sector median, growth-adjusted
FAIL
P/E 31.6x · rev +46.8%
Fair-Value Bridge method blend, measured
Growth/quality-adjusted PER (w20%)₩1,416,261
Analyst target (38) (w80%)₩2,712,489
Owl fair value / share₩2,453,243
Upside at ₩1,399,000+75.4%
MODEL INPUTS (DISCLOSED)
P/E · P/B31.6x · 11.26x
ROE (latest FY)35.6%
ROE (3-yr avg, normalized)15.1%
Revenue CAGR (2-yr)+72.2%
Band: CHEAP · justified-multiple model with quality/growth adjustments — an estimate, not a promise.
Sensitivity value per share
EARNINGS ↓ / MULTIPLE →
-10%
base
+10%
-10%
₩1,147,171
₩1,274,635
₩1,402,098
base (measured)
₩1,274,635
₩1,416,261
₩1,557,887
+10%
₩1,402,098
₩1,557,887
₩1,713,675
Simple sensitivity (reference): ±10% arithmetic on the measured P/E-method value — earnings basis × fair multiple. Not a forecast; no projected figures are used.
Read this like a memo: the model says ₩2,453,243 and the Street consensus says ₩2,712,489 (38 analysts). The gap is the assumption set — both are computed from disclosed inputs above; underwrite the difference yourself.
Information only — not investment advice · Sources: exchange data · SEC EDGAR · DART · Owl Labs models
No tracked manager reports a US-listed long position in 000660 this quarter. 13F covers only US-listed instruments (e.g., ADRs) — "not held" may simply mean not reportable.
13F filings disclose US-listed long positions ~45 days after quarter end. "Style lens" lines relate this stock to each investor's public style using measured metrics only — they are not the managers' actual views. For Korean stocks, 13F covers only US-listed instruments (e.g., ADRs) — "not held" may simply mean not reportable.
Information only — not investment advice · Sources: exchange data · SEC EDGAR · DART · Owl Labs models
Measured state: RSI 36.1 (neutral), MACD below its signal, no MA cross is active. Read the tape as one input, not a verdict.
Catalyst Calendar
No confirmed earnings date in calendar
Every weekday — quant screening refreshscore update
Quarterly — 13F filing window (~45d after quarter end)13F refresh
Forward Signals (unscored) consensus data · never scored
Street target consensus₩2,712,489 (38)
Implied upside vs price+93.9%
Next-quarter EPS estimate—
Next-quarter revenue estimate—
Shown for context, never scored: the composite grades only reported fundamentals. Forward numbers are third-party consensus — treat them as the market’s assumption set, not ours.
Risk Register
MID
유상증자(지분 희석)(disclosure flag · DART)
HIGH
Drawdown state. Price sits -53.2% below the 52-week high (42th percentile of the range) — measured tape, not a judgment of cause.
WHAT THE BEARS MUST PROVE / WHAT THE BULLS MUST PROVE
Bears: that the measured quality base cracks — ROE 35.6% and op. margin 48.6% are what the score leans on. Bulls: less on price — the model's ₩2,453,243 sits +75.4% above; they must show the measured fundamentals hold through the next filings.
Information only — not investment advice · Sources: exchange data · SEC EDGAR · DART · Owl Labs models
Source: company filings (annual). Only reported fiscal years are shown — no forward estimates. ROE 35.6% · Net margin 44.2%.
News Digest RSS · external media
07/27 — SK Hynix ADR Falls Below IPO Price to Record Low After US Listing (bloomingbit)
07/27 — South Korea’s Leveraged ETF Gambit Backfires as SK Hynix and Samsung Sink (Barron's)
07/27 — Secondary Batteries to Lead After Semiconductor Surge (조선일보)
07/27 — Nvidia in Talks to Finance OpenAI, Report Says. What It Means for the Stock. (Barron's)
External media — accuracy rests with each outlet.
Methodology & how to use
Every number in this report is measured — pulled from the screener database, SEC EDGAR 13F filings, DART disclosures, company financials, or deterministic models (justified-multiple fair value). Sentences are rule-based templates bound to those numbers; nothing is generated, so nothing is hallucinated. Axis weights and the sector adjustment are proprietary; axis names, raw scores, and every input above are disclosed for audit.
Known limits, stated plainly. ① Backward-looking — the score grades reported fundamentals and today's price; future inflections (a cycle turn, next year's earnings) are not scored and appear only as unscored context (street target, catalysts). A low score can mean "great business, full price". ② The composite is this framework's ranking, not an absolute quality of the company. ③ 13F covers US-listed holdings only. Use the report as a screener; pair it with your own forward and qualitative work before acting.
B
BALANCED CASE
Final judgment. 000660 scores 70/100 (grade B) — rank #159 of 1036 in the KR universe. Axis split: quality 40/40 · value 33/80 · attention 1/20. Model fair value ₩2,453,243 (+75.4% vs price). The verdict changes only when the data changes.
Grade rubric: A 85+ (top conviction basket) · B 65+ (sound, price matters) · C 45+ (no rush) · D below (weak case). The verdict changes only when the data changes — next screening runs every weekday.
Owl Labs is an online information provider. This document is compiled from public data sources (exchange quotes, SEC EDGAR 13F, DART filings, company financials, news RSS) and Owl Labs quantitative models. It is not investment advice, not a solicitation to trade any security, and past data does not guarantee future results. Figures may differ from real-time values. Final investment decisions and responsibility rest solely with the investor. Unauthorized redistribution prohibited.
