Ideas
Korean biotech lacks supply; avoid Alteogen.
Korean biotech sentiment is poor because the large speculative funds that used to drive the sector have been hurt by earlier losses. Alteogen's fundamentals are not bad, but supply is not attaching; if the 250,000 won support breaks, it could struggle further until institutions or foreigners return.
AI speed-control noise; stay long AI infrastructure.
The 'AI speed control' debate is noise, not a real demand signal. AI infrastructure is still the market's main growth engine, and investors should stay exposed to semiconductors and related AI hardware rather than treating the debate as a reason to exit.
CPO optical theme broadening; long related names.
CPO and optical communication is a broadening AI data-center theme as copper is increasingly replaced by optical links. Seongho Electronics has CPO inspection equipment exposure, Daehan Optical Telecom and RFHIC and RF Materials are domestic communication names, and Lumentum and Coherent are overseas reference plays.
Peptron uncertainty; avoid ambiguous biotech.
Peptron has been delayed and vague on its second plant and technology-evaluation contract deadline. Investors should avoid such ambiguous biotech names until there is a clear result.
Power equipment benefits from AI buildout.
AI hardware strength should keep flowing into power equipment and wires. LS Electric, Hyosung Heavy Industries, HD Hyundai Electric, Sanil Electric, Iljin Electric and LS are beneficiaries as AI data centers require more electrical infrastructure, with rotation into laggards possible.
SK hynix ADR premium supports local upside.
SK hynix's ADR premium is a useful valuation guide. The premium had spiked to 45% and has narrowed to around 30-35%; at a 30% premium, the local shares should rise another roughly 5% toward the high-1.9 million won area. Memory and HBM demand remain strong.
Samsung Electronics undervalued; watch HBM catch-up.
Samsung Electronics is relatively undervalued versus SK hynix and could catch up through HBM base-die and Tesla-related opportunities, but the stock has not moved and investors are impatient. It is a catch-up watch rather than a fast momentum trade.
Data-center rents rising; long neoclouds and hyperscalers.
Data-center rental prices are rising 20-25%, showing tight demand. Nebius has less debt and a simpler customer-funded model, while CoreWeave has more debt and noise; both are earning well, and the tight data-center market also benefits hyperscalers like Amazon and Alphabet.
Daehan Electric Wire overextended; trim exposure.
Daehan Electric Wire has run up sharply and is trading above 40x forward earnings. The chart has no upper tail, but further upside would make valuation burdensome; holders can consider trimming into Monday rather than chasing.
GNC Energy overvalued; avoid.
GNC Energy's backup-generator theme is interesting, but the stock has already rallied and trades around 13x with only mid-teens growth. The risk-reward has become uncomfortable, so it is not an attractive entry.
Doosan bottoming; hold for CCL growth.
Doosan is investable around current levels as it invests 1 trillion won in CCL and PCB materials and completes the SK Siltron acquisition. The stock has been heavy, and valuation looks less burdensome when looking through the holding-company structure, so holders can stay with it.
Doosan Tesna valuation reset; watch rebound.
Doosan Tesna has fallen to a more reasonable valuation after target-price cuts, and its post-processing exposure to Samsung image sensors and Tesla's Taylor project gives it rebound potential once expectations reset. It is a watch-list rebound idea, not an immediate earnings catalyst.
Oil is key macro risk; watch.
Oil is the last major macro risk. If oil stabilizes or falls, many market concerns ease; if it stays high long enough, it can feed inflation expectations and force more tightening. Watch crude and diesel prices closely.
NVIDIA leads AI semiconductor ecosystem.
AI semiconductor demand remains intact. Jensen Huang said Nvidia chip sales could double next year, and memory bottlenecks are likely to worsen, supporting the GPU-led AI semiconductor ecosystem even as memory costs become a 2027 risk.
Power bottleneck; long on-site power.
Power is the physical bottleneck for AI data centers. On-site power solutions such as backup generators, fuel cells and LNG will be needed first because large-scale nuclear and SMR projects take too long, so power infrastructure should stay strong.
Generac wins Amazon backup generator deal.
Amazon signed a backup-generator supply contract with Generac and received warrants, echoing Nvidia's optical investments. The deal shows how desperate data centers are for power and should keep Generac and related power-generation names in focus.
Semiconductor equipment benefiting from capacity expansion.
Semiconductor capacity expansion is confirmed by AI demand, which should benefit equipment and materials. Back-end packaging is especially attractive; PSK Holdings has hit new highs and Wonik IPS has rebounded sharply, though investors should still watch valuations and rotation.
Cosmetics ODM leaders; avoid third-tier names.
In cosmetics, ODM leaders such as Kolmar Korea and Cosmecca Korea are stronger than third-tier names and have held up better during sector pullbacks. However, when even laggards rally together, that is a signal to be cautious on the broader cosmetics trade.
Optical communication crowded; rotate to batteries/nuclear.
Optical communication became crowded as money rotated into Daehan Optical Telecom, RFHIC, KMW and OI Solution. With low market turnover, the sector needs fresh inflows to keep rising; otherwise, selling into strength and rotating to laggards is the better strategy.
Secondary batteries benefit from AI power needs.
Secondary batteries and ESS are part of the AI value chain as data centers need power stabilization and backup. They have lagged while optical communication rallied, creating a potential rotation trade into batteries and nuclear.
Construction benefits from data-center and nuclear buildout.
Data centers and nuclear plants must be built, which ultimately benefits construction companies. Hyundai E&C and Samsung E&A are ways to play the AI data-center and power-plant construction cycle, especially after recent pullbacks.
Doosan Enerbility benefits from nuclear power demand.
Nuclear power is needed for the AI data-center buildout. Doosan Enerbility can participate in both the Westinghouse and Team Korea models, and Westinghouse's equity investment reduces the intellectual-property risk that previously blocked Korean nuclear exports.
Power transformers benefit from AI data centers.
Power transformers and grid equipment remain core AI-infrastructure holdings. The group has lagged during the recent wire rally, and rotation could favor names such as Hyosung Heavy Industries, HD Hyundai Electric and LS Electric.
Sanil Electric gains from transformer market entry.
Sanil Electric is entering the large-transformer market and its stock is already stronger than Hyosung Heavy and HD Hyundai Electric. If transformers run, Sanil Electric may have more upside because it is earlier in its re-rating.
KOSPI dips are buy opportunities next week.
Next week is a buy-on-dips week for KOSPI. If the market weakens Monday-Wednesday before Chuseok, investors should buy and look for a move toward the 7,200-7,500 area.
SK hynix rebound to 2.1 million won.
SK hynix should see a technical rebound toward 2.1 million won, with the timing roughly coinciding with the end of its buyback. If the stock falls before then, buy; if already holding, hold.
AI speed control unrealistic; own AI infrastructure.
The AI speed-control debate is not realistic because companies cannot credibly open their models and investment plans to rivals. AI capability is still advancing, so investors should focus on the expanding AI value chain rather than the noise.
Power infrastructure is next AI bottleneck.
Power infrastructure is the bowl that holds AI. Data centers are being delayed because they lack energy, so transformers, wires, fuel cells, on-site generation and later nuclear, SMR and renewables are necessary investments over the next several years.
6G communication infrastructure needed for AI robots.
Communication infrastructure is the missing piece in AI infrastructure. Current 5G speeds are insufficient for autonomous driving and humanoid robots; 6G and broader communication infrastructure will need substantial investment.
Samsung dividend may reveal foreign interest.
Samsung Electronics' 30 trillion won dividend is a near-term foreign-supply test. If foreigners buy before the late-September dividend date, it signals real interest; if they do not buy despite the payout, it signals little interest in the stock.
Hyosung Heavy offers decade-long power growth.
For a 10-year investment, Cha would buy Hyosung Heavy Industries because power infrastructure demand will grow as data centers consume huge amounts of electricity. Hyosung has global transformer share and can benefit from the long power buildout despite competition from global peers.
ESS may be next AI power trade.
ESS and secondary batteries may be the next AI power trade as renewables and data centers need storage and stabilization. Cha frames this as a possible later-stage AI momentum shift rather than an immediate recommendation.
This 3PRO TV (삼프로TV) video, published September 18, 2026,
features Lee Kwon-hee, Kim Jang-yeol, Jang Woo-jin, Hwang Yoo-hyun, Cha Young-joo
discussing 196170.KQ, Korean Biotech, AIQ, SMH, 081580.KQ, 010170.KQ, 218410.KQ, 327260.KQ, LITE, COHR, 087010.KQ, 010120.KS, 298040.KS, 267260.KS, 062040.KS, Iljin Electric, LS, 000660.KS, 005930.KS, NEBIUS, CoreWeave, AMZN, GOOG, 001440.KS, 119850.KQ, 000150.KS, 131970.KQ, HO=F, WTI, NVDA, Power infrastructure, LNG, Fuel cells, GNRC, 067310.KQ, WONIK IPS, Korean semiconductor equipment, 161890.KS, Cosmecca Korea, Optical Communication, 032500.KQ, OI Solution, Korean battery/ess sector, Hyundai E&C, 028050.KS, 034020.KS, URA, EWY, Transformers, NLR, Communication infrastructure, 6G.
32 trade ideas extracted by AI with direction and confidence scoring.
Speakers:
Lee Kwon-hee,
Kim Jang-yeol,
Jang Woo-jin,
Hwang Yoo-hyun,
Cha Young-joo
· Tickers:
196170.KQ,
Korean Biotech,
AIQ,
SMH,
081580.KQ,
010170.KQ,
218410.KQ,
327260.KQ,
LITE,
COHR,
087010.KQ,
010120.KS,
298040.KS,
267260.KS,
062040.KS,
Iljin Electric,
LS,
000660.KS,
005930.KS,
NEBIUS,
CoreWeave,
AMZN,
GOOG,
001440.KS,
119850.KQ,
000150.KS,
131970.KQ,
HO=F,
WTI,
NVDA,
Power infrastructure,
LNG,
Fuel cells,
GNRC,
067310.KQ,
WONIK IPS,
Korean semiconductor equipment,
161890.KS,
Cosmecca Korea,
Optical Communication,
032500.KQ,
OI Solution,
Korean battery/ess sector,
Hyundai E&C,
028050.KS,
034020.KS,
URA,
EWY,
Transformers,
NLR,
Communication infrastructure,
6G
