Video summary

AI 랠리 끝난 걸까? 지금은 현금 가지고 기다려야 하는 이유 | 21년 프랍트레이더 김진

Main summary

Key takeaways

Finance

Finance-specific summary (markets/investing focus)

Big picture (what changed)

  • Core thesis: In the U.S. AI hardware leadership group, the “stable correction range” broke, and the correction widened starting Thursday night (U.S. time / Korea time).
  • Implication: This typically suggests the consolidation/stagnation period may last longer and the odds of “leading stocks changing” increases (i.e., leadership may rotate to different names within AI/semis/adjacent hardware).
  • Scope: The narrator argues the AI rally isn’t confirmed to be “fully over” yet—weakness appears in parts, not necessarily across all AI hardware.

U.S. market: semiconductor-led weakness + leadership risk

Semiconductor sector (overall read)

  • Semiconductor index: Broke below the lower bound of a stabilization/range from Thursday through Friday, interpreted as deteriorating stability.
  • Takeaway: Semiconductor “stability adjustment” is worsening. It may not be the end of the entire AI trend, but the control/consolidation could extend longer and worsen.

CPU / chip design names (examples mentioned)

  • AMD: Maintains a stable/sideways range (described as still “holding ground”).
  • Intel: Previously showed stability, but described as deviating starting Thursday.
  • ARM: Rose on Friday, but is described as having a strong downward trend after breaking out of its stable correction.

Semiconductor equipment (capital goods for chips)

  • Applied Materials (AMAT): Broke below prior lows / stable-correction support.
  • Lam Research (LRCX): Broke below previous lows; weakness interpreted as beginning to show “warning-type trends.”
  • Longer-term frame: Even if it’s not a full “breakout failure” across every chart perspective, equipment is viewed as showing more damaging deterioration versus earlier.

Memory / Korea-relevant linkage

  • Micron Technology: “Loss of control” observed, but a complete breakout from the entire rally is not yet confirmed.
  • SanDisk (likely referring to WD): Weaker trend more pronounced from Thursday; broke below stable correction and stayed weak into Friday.
  • Narrator’s structure: Memory + equipment show some areas with breakout/weakness, but not necessarily universal collapse across the whole AI complex.

“AI infrastructure” but non-core semis (network/power/cooling)

  • Dell (network hardware):
    • Main issue: Wednesday sharp drop
    • Then Thursday weakness
    • Friday no further deterioration → interpreted as stable correction maintaining
  • Coherent: Breakaway from the uptrend; risk intensifying short-term.
  • Dari / Cisco (name unclear for “Dari”): Presented as not deteriorating to a full breakdown of stable correction (more neutral).
  • Generators / power / cooling equipment:
    • Caterpillar (CAT): Slight risk/weakness; narrator suggests it’s not clearly a definite “trend ended” call yet.
    • Other industrial names referenced include a garbled “number bar” (no clear ticker provided).

Big Tech + software: why Nasdaq stability doesn’t equal “AI hardware is safe”

Nasdaq stabilization explanation

  • Volatility from AI hardware corrections is offset short-term because Big Tech / software were relatively stronger.

Stocks cited

  • Apple: Rally, but narrator says it’s not crucial for AI investment (least relevant among Big Tech for AI).
  • Microsoft, Amazon, Alphabet: Fluctuations described as non-trending / not showing strong positive AI-led momentum.

Software enterprise “barometer” (agentic AI spread)

  • CrowdStrike and Snowflake are cited.
  • Narrator’s point: they don’t show a special upward rally, but that still doesn’t mean the upward AI trend is fully gone.
  • Conclusion: Enterprise data software flow still matters for whether AI agent adoption sustains the rally.

Key performance/risk recommendations (explicit strategy)

Main recommendation: Increase cash weighting

The narrator explicitly recommends raising cash proportion due to:

  • Widening correction range in leadership stocks
  • Higher uncertainty about which stocks will lead next

Rationale:

  • Buying “cheap” is not the same as buying the next leaders.
  • If leadership rotates, cash provides optionality to buy new leaders later.

Explicit caution:

  • Warning against a mindset like “I can’t sell now” (binary behavior).
  • Holding through a crisis can ruin the next opportunity if leadership changes.

Korea/Asia transmission (why the domestic market matters)

Lead-lag timing

  • Korea broke out of stabilization earlier (~1 week to 1.5 weeks ahead) than the U.S.
  • As cross-checking evidence, Asia weakness is described as confirming the Thursday night U.S. changes, focusing on:
    • Korea: KOSPI 200, KOSDAQ
    • Japan / Taiwan: Taiwan-weighted index mentioned

What’s blamed as the transmission center

  • Semiconductors (e.g., Samsung Electronics, SK Hynix noted as remaining “leading” names).

Domestic market summary (last week)

  • KOSPI 200: Sharp drop/rebound cycles with renewed decline—downtrend maintained, not clearly reversed.
  • KOSDAQ: Downtrend earlier; last week described as not worsening dramatically, but still failing to break out of the downtrend channel.
  • Banks: Relatively strong; narrator suggests banks can soften deterioration, but warns against overly optimistic interpretation.

Debate/disclaimer-style theme: leverage vs fundamentals (KOSDAQ)

  • Narrator addresses claims that KOSDAQ was “ruined” due to leverage (Neverage/leverage launch).
  • Counterpoint:
    • KOSDAQ’s sector composition (garbled, but includes):
      • secondary batteries
      • pharmaceuticals / biotechnology
      • semiconductor materials/parts/equipment
    • Underperformance is attributed more to sector weakness than leverage alone.
  • Examples mentioned:
    • LG Energy Solution, Samsung SDI (secondary batteries)
    • Samsung Biologics, Yuhan Corporation (biopharma/biotech)

Timelines emphasized

  • Thursday night (U.S. time): Leadership/stability broke in the U.S. AI hardware group.
  • Thursday → Friday: Semiconductor stabilization deterioration continued.
  • Korea vs global lead-lag: Korea moved earlier; domestic impact with delay of about 1 week to 1.5 weeks.
  • This week outlook: Repeated message: no one knows the next move (compared to a coin toss), so strategy should reflect uncertainty and longer consolidation risk.

Behavioral guidance: reduce/sell vs uncertainty

If holdings were already cut significantly

  • Don’t treat early-week drops as an automatic trigger to sell everything even more—may require incremental actions.

If holdings weren’t reduced yet

  • There may be a timing point to reduce appropriately, but exact timing is unknown.

Disclosures / disclaimers

  • No explicit “financial advice” wording appears in the provided subtitles.
  • Advice is framed as investment philosophy, with an emphasis on uncertainty.

Instruments / tickers / indices mentioned

U.S. / Global

  • AMD
  • Intel
  • ARM
  • Applied Materials (AMAT)
  • Lam Research (LRCX)
  • Micron Technology (MU)
  • SanDisk (ticker not clearly stated; typically related to WD)
  • NVIDIA (NVDA)
  • Broadcom (AVGO)
  • Dell
  • Coherent
  • Cisco
  • Caterpillar (CAT)
  • Apple
  • Microsoft
  • Amazon
  • Alphabet
  • CrowdStrike (CRWD) (implied by “Crowdsoft”)
  • Snowflake (SNOW)

Korea / Asia

  • KOSPI 200
  • KOSDAQ
  • Samsung Electronics
  • SK Hynix
  • Samsung Electro-Mechanics
  • LG Energy Solution
  • Samsung SDI
  • Samsung Biologics
  • Yuhan Corporation
  • Taiwan Weighted Index

Methodology / framework (as described)

  • Chart/regime framework (qualitative):
    • Identify whether leadership groups are in a “stable correction range” versus breaking below the lower bound (stability deterioration).
    • If the correction range widens, infer longer stagnation and higher leadership rotation risk.
  • Cross-sector confirmation:
    • Compare semis within sub-sectors (CPUs, equipment, memory) to locate where breakdown occurs.
    • Check adjacent AI infrastructure (network/power/cooling) to judge whether the AI complex weakens broadly or only in parts.
  • Optionality framework:
    • If leadership rotation risk rises → increase cash weighting rather than trying to buy “cheap leaders” blindly.
  • Behavioral risk framework:
    • Avoid binary thinking (“can’t sell”) during regime shifts to preserve the ability to act when leadership rotates.

Presenters / sources

  • Presenter: Kim Jin (also referred to as Jin-saem Kim Jin)
  • Channel/host reference: Kim Jin’s Investment Camp

Original video