Kim Dante, Lee Hyo-seok and Kim Kwang-seok differ on how far the market correction has progressed and compare their reasoning rather than predict an exact turning point. Kim Dante interprets stalled 3rd-party estimates of Anthropic annualized revenue and the spread of inexpensive Chinese open-weight models as pressure on frontier AI companies’ pricing and growth. He cites model-routing practices at startups around him and shares on OpenRouter, which are not treated as definitive statistics for the entire AI market. Lee Hyo-seok distinguishes slower revenue growth from stalling and suggests competition from OpenAI as another possible cause alongside open-weight models. He then explains the chain from model-company revenue through compute, data centers and memory spending, and the leverage used by business operators. He outlines possible recovery in confidence through earnings, capital injections or equity restructuring, without guaranteeing NVIDIA support or a market bottom. The video ends as a further counterargument begins, so that argument cannot be established from this installment alone.

Economics Reader (Kim Kwang-seok TV)

How far has the correction progressed? Three different assessments

Reading estimated Anthropic annualized revenue as a signal for the AI investment cycle

Low-cost open-weight models, routing and the scope of platform market shares

Growth slowing versus stalling: from model-company revenue to memory investment

Leverage and credit support: earnings, capital injections and equity restructuring scenarios

The KOSPI correction is not over yet…Chinese AI is changing the game | Let’s debate with the Economics Reader | 3-way discussion: Lee Hyo-seok, Kim Dante and Kim Kwang-seok [Part 1]

The video introduces a Tesla Roadster demonstration, describing the striking presentation of downforce and gas-propulsion technology. The speaker praises its technological significance while identifying uncertainties around ordinary drivers' safety, road use, pricing and production timing. He contrasts skepticism about the small contribution of a low-volume expensive vehicle with the potential diffusion of some technologies. In the second half, he connects the prominence of SpaceX at the event with an interpretation of it as marketing for a merger rationale. He then treats business synergies from vertical integration and the possibility of stronger Musk control as separate questions. Launch dates, merger schedules, exchange ratios, ownership and voting-power figures in the video are forecasts or assumptions introduced by the speaker, rather than independently verified transaction terms in this note. All directional outlooks are the speaker's conditional judgments and do not establish certain short-term price predictions.

Syuka World

The secret behind Tesla's flying car

October (10) 20 Indicator Tracker said institutional selling beat foreign spot selling of 1 trillion won, called weak flows a symptom, and cited a 230 of 250 days reflection rule with a 48 to 52 mix. With foreign selling worth only 7 thousand-hundred-million units and CoreWeave down 25 percent, he framed an AI retreat with speculation exits, split AI from AI stocks, and flagged individuals buying 1 trillion won. He attacked 55 valueless minutes of 1 hour, warned against news-driven trades, cited a minus 40 percent tuition lesson favoring index buys, and urged watching France irrelevance, CoreWeave unsecured paper, and CDS cracks. Foreign weight at 38 percent, KOSPI closing at 5,552 trapped near 5,500, down odds of 70 versus 30, KOSDAQ sliding from 781 toward 666 buys, Nasdaq on its 60-day line, and bizarre won strength were wrapped up. US 10-year yields at 5.11 with a 20-year auction tomorrow, WTI rolls near 101 dollars, Samsung corrected to 20man 9,500 won with buys from 19man to 15man won, Hynix down 5 percent, and dollar-yen at 198 were reported. He rejected 4,500, saw 8 percent room with a broken 60-week line, and laid out a 5,000-touch then 6,000-rip assumption on a homemade 21-day average risking 3 to 4 percent for 10 to 20 percent. KOSDAQ 666 with 11 percent room toward 1,200, Hanwha down 11 and 21 percent, EcoPro BM stop advice at 90man won, WTI toward 110, gold at 3,927, weekly 5,120 levels, 20 percent cash, and a 10 percent mental stop guarding 90 percent were wrapped up. All directional views are the speakers' conditional judgments and do not confirm any short-term price forecast.

Alsangmu

Sorry to those stuck in Korean stocks [Indicator Tracker 1020]

On Chesley TV's Womaesinbak, Director Park Se-ik prescribes contrarian trading for the box range. He reflects on missing the 9,200 top, margin-fueled false demand, and managers stuck underwater. With margin balances above 30 trillion won, the strategy is buying dips and selling rallies. Despite DRAM gains and Samsung Electronics profit growth ahead, stocks face macro-driven de-rating, with a 19 ten-thousand-won support line. He argues immigration beyond the threshold cuts wages and breeds conflict and unrest, favoring dollar hegemony. He relays JP Morgan's no-certain-bottom call and warns against blind faith in brokerage reports. All directional outlooks are the speaker's conditional judgments, not definitive short-term price predictions.

ChesleyTV

Margin balances still at 30 trillion won: contrarian trades for the box range! | Chesley Investment Advisory Director Park Se-ik [Womaesinbak / 26.10.20 Tue]

On Chesley TV's Womae Sinbak, Director Park Se-ik reads the newspapers into a selloff playbook in a 58-minute show. He frames inflection-point volatility as a tide-and-monsoon-front tug of war, reads widening swings as a nearing bottom, and urges challenge over self-blame. Citing KB Securities on Samsung, he eases profit fears with 80% of next-year volume on long-term contracts, HBM4 100% contracted, and a 555 trillion won next-year operating-profit outlook, eyeing 19 man won as support. Europe shows manufacturing collapse and Chinese-car gains with an 80 man French protest, 2,400 arrests, spreading immigrant-jobs strife, and far-right votes up from 10% to over 23%, he reports. US AI-infra cracks widen power, permit, and funding strains, and he compares the 2025-2032 spending estimate of 10 trillion 3 thousand hundred-million dollars at 3.63% of GDP with rail and telecom booms. He walks Samsung's chip history from Chairman Lee's 1974 buy at age 32 of a 50% stake to 1992 world No.1 in DRAM, plus today's 40% DRAM, 30% NAND, and 32% HBM shares. JPMorgan sees extreme undervaluation at 4 times forward PER and near 1 times PBR yet no flow proof of a bottom, naming memory and hyperscaler prints, calmer volatility, and the midterms as catalysts, he reports. All directional views are the speaker's conditional judgments and do not settle any near-term price forecast.

Chesley TV

JPMorgan: KOSPI very cheap but the bottom is uncertain | Director Park Se-ik of Chesley Investment Advisory [Womae Sinbak / 26.10.20 Tue]

An October (month 10) 20 Yeouido Insight diagnosing the foreign-flow-driven market with the Geumsigong CEO. It treats yesterday's bounce as mechanical, flagging the WTI break above 100 dollars and French bond-dump fears among rate drivers. It rebuts as exaggerated three semiconductor rumors on Intel LPDDR, China AI self-reliance, and MVHBM. It explains AI skepticism via the CIO 86 percent survey and data-center funding, power, and approval bottlenecks. It centers on Oracle's BBB-minus credit risk, OpenAI's share of backlog, and holes in takeover optimism. It also covers Hynix weakness, the Solidigm duplicate-listing debate, Middle East escalation, and the November (month 11) 3 midterm schedule. It proposes watching next week's big-tech earnings, raising cash, and reading the stabilization fund as a bottom signal. All directional views are the speaker's conditional judgments and do not confirm short-term price forecasts.

SamproTV 3PROTV

100-dollar oil shakes AI investment again... why Middle East escalation even stirs rates | Hong Seon-ae, Jang Woo-jin, Geumsigong CEO [Yeouido Insight]

An October (month 10) 20 Korea close briefing covering the exchange plunge, foreign selling, and spreading decliners. It reviews the morning Hynix plunge and Solidigm listing controversy with the afternoon broad fall and KOSDAQ weakness. It centers on the credit-balance burden, spreading forced selling, and the stabilization-fund deployment debate. It also covers the Solidigm shareholder-versus-growth debate and OpenAI fundraising plus NPS, rare-earth, and KEPCO news. It critically examines the Google CIO versus Jefferies AI investment debate and exaggeration in the securities-staff rule-violation coverage. The speaker sees Korean stocks as cheap with bottoming sentiment and urges holding on. Figures, titles, and proper nouns follow the captions without asserting uncertain transcriptions. All directional views are the speaker's conditional judgments and do not confirm short-term price forecasts.

JeungsigakdogiTV

[October (month 10) 20, Tuesday Korea market] another plunge! No one stepping in to buy! | stabilization fund, please use it now | Hynix treasury buying over 98% done | 3000-eok-won forced selling over 5 days?

Shuka brought Cathie Wood back after 4 months, asking about Samsung and Hynix plunge fears, and she answered the whole semiconductor space sits early in a multiyear rise past a big correction, welcoming market fear. Skipping direct memory bets over volatility, Chinese entry, and hyperscaler weight cuts, while shunning commoditized turf and warning of margin pressure, she still foresaw robotaxi and humanoid demand at the margin with an HBM shift to higher value, urging buy-the-dip since AI is only round 1 of 9, still at the parking lot. September (9) 9 buying of Meta for 279 90man dollars as transcribed, about 380 in hundred-million-won units, against selling 2,780man dollars of Google were small moves to fund beaten-down buys, while Muse at 200man users a day, number 1 for 28 straight days, untroubled FCF-burning aggression, advertising holy grail hopes, USDC networks, online commerce expansion, and 50 percent were wrapped up. The 2030 bull case of 150man dollars against 8man 4,000 dollars now, nearly 15 times, rested on digital gold, a 0.1 correlation, and 20 million of 21 million units mined, with base 75man and stablecoins at 3,000 in hundred-million units versus 3 trillion in crypto, plus ARK Venture Fund tokenization. Hinting at SpaceX engines for the Roadster, she backed a Tesla-SpaceX merger, citing vertical integration, Terafab ASIC specials for space data centers, Moon and Mars humanoids, and Elon-TSMC Terafab talks. Past US 10-year yields over 5 percent toward 5.3 percent, near a 20-year high, she still called 5 to 6 percent normal and the past 30 to 40 years the low-rate anomaly, citing nominal GDP of 4.8 against 6 to 7 percent ahead, unlike the 2022 Fed shock of 0.25 to 5.5, expecting rate-up outperformance as in 2017 and 2018. Averaging into space, buying Tesla on dips, gaining from Rocket Lab on the Falcon 9 to Starship shift and neutron delays, and favoring Korean LG energy storage, semiconductors, and Han Aerospace were reported, with EV margins of 15 to 16 percent against robotaxi at 80 percent plus, a 5-year subscription shift, harmless midterm gridlock, Trump policies to 2029, late 28 elections, and falling oil. All directional views are the speakers' conditional judgments and do not confirm any short-term price forecast.

PLUS TV

SpaceX engines on the Tesla Roadster? The merger synergy Cathie Wood expects from the two companies

The speaker interprets Korea's greater vulnerability through differences in US and Korean market declines, trading hours and market weights. He describes reactions to war-related news and changes in oil prices and yields, while saying investors cannot control or predict the timing of a ceasefire or escalation. He does not treat US after-hours rebounds as strong evidence of a trend reversal and notes the gap between Korea's two main semiconductor shares and other stocks. While allowing for a rebound that day, he also assumes another decline is possible and presents his own short-term plan to sell part of his holdings and reassess late in the session. He considers valuation alone insufficient to confirm a bottom and wants price support and a period of stability. He explains that abandoning a loss-cutting plan because of a rebound, followed by renewed declines, can deepen distrust of one's judgment as well as financial losses. Percentage moves and trading plans in the video are observations, forecasts and personal responses at the time, rather than verified subsequent outcomes or instructions applicable to all investors. All directional outlooks are the speaker's conditional judgments and do not establish certain short-term price predictions.

R. Sangmoo

Doing nothing is also an option. If asked whether the expected return now is high enough relative to the risks involved, I suspect few people could answer easily. I think this is a time to consider several possibilities from multiple angles, rather than settle on a single direction.

US–Korean market divergence: timing, weight and attention

War news, oil and yields: major effects and unpredictable timing

Limits of after-hours rebounds, semiconductor concentration and response conditions

A rebound is not a confirmed bottom: price support and short-term responses

The psychology of reversing a loss-cutting plan and regaining control

Daily strategy (1021 Wed)

Compare three views on the correction and distinguish the links between AI revenue, low-cost models, compute contracts and leverage.

The KOSPI correction is not over yet…Chinese AI is changing the game | Let’s debate with the Economics Reader | 3-way discussion: Lee Hyo-seok, Kim Dante and Kim Kwang-seok [Part 1]

Timeline

Was this helpful?

Questions & opinions