R14 strike-window staging — cohort 7D red is profit-taking, NOT a demand-side break; publish-gate still held

Update

R14 strike-window staging — cohort 7D red is profit-taking, NOT a demand-side break; publish-gate still held

  • Type: refresh (subtype: continuity-annotation)
  • event_id: 2026-05-19-memory-supercycle-r14-strike-window-staging
  • Source: R14 next-day research desk fanout — workflow/RUNS.md 2026-05-19 r14-next-day-research-desk-fanout. Inputs: published/research-notes/2026-05-19-r14-memory-supercycle-strike-window-staging.md; published/scan-result-buckets/2026-05-19-r14-next-day-research-desk-fanout.json; targeted web verification of strike day 0 status + cohort 7D catalyst attribution + TrendForce 2Q26 contract-price commentary + analyst house-view shifts (Korea Herald 5/19, Korea Times 5/19, 247WallSt 5/12, Motley Fool 5/18, TrendForce 3/31, Tom's Hardware MS-DRAM-unaffordable).
  • Decision context: R10 perspective stand-up gated publish on "cohort breaks green." 36 hours before Samsung NSEU strike day 0 (2026-05-21), the cohort is unanimously RED 7D — MU -14.3%, SNDK -13.9%, STX -11.2%, WDC -11.1%, HY9H.F -4.6%. R14 asks: did the cohort already price the strike, or is the de-rate breaking the supercycle thesis?
  • Verdict — the 7D red is profit-taking + recycled bear-narrative cover, NOT a demand-side de-rate. The thesis is intact and sharpening; the publish-gate stays "cohort breaks green" and is now harder to clear, which is itself the honest read. Three confirmations:
    1. No fundamentals catalyst broke in the 7 days. TrendForce 2Q26 forecast (DRAM +58-63% QoQ / NAND +70-75% QoQ) has NOT been revised; no hyperscaler capex cut; no cohort earnings miss or guide cut; no analyst house-view shift specifically this week (Morgan Stanley DRAM-unaffordable + Goldman MU-Neutral are Q1 notes being re-cited as cover).
    2. The 5/12 sector unwind was explicitly profit-taking. 247WallSt and Invezz both characterized MU -9% / WDC -8% / SNDK -9% as "sector-level profit-taking rather than any company-specific deterioration" against YTD-into-5/12 gains of MU +179%, WDC +200%, SNDK +552%.
    3. The 5/18 follow-through was attributed to "strike + China + valuation" — three different cover stories. When the cover-story stack thickens, the unwind is price-action-led with rationalization downstream.
  • Strike day 0 status (2026-05-19): still on track. Korea Herald / Korea Times 2026-05-19 — last-ditch NLRC mediation in session today; both sides "still at odds" but "both making concessions"; union has stated it will not negotiate further until after the strike if today fails. Fortune 5/17 — consensus impact ~$700M/day, removing ~3-4% global DRAM / 2-3% NAND over 18 days.
  • Strange tell: the strike-bid did not arrive. Naive supply-shock logic says non-Samsung memory names should be bid 36 hours before a confirmed ~3-4% global DRAM removal. They are not — the market is pricing this as risk-off (uncertainty + demand-risk), not scarcity-on. The interpretation is that smart money is selling INTO the strike, not buying through it — which says they are de-risking the YTD-runner positions, not abandoning the cycle.
  • Effect on thesis: HOLD — posture unchanged. Cohort = MU/SNDK/STX/WDC/HY9H.F kept. Status active, priority high. Publish-gate "cohort breaks green" stays. Note for sharpening at the next refresh: the gate should be tightened to an operational definition (e.g., "5/10/14-day cohort price ROC turning positive with RSI flooring at ≥40 across at least 4 of 5 names") to avoid future ambiguity.
  • Unresolved checks (→ SCANS/RESEARCH): (a) TrendForce / DRAMeXchange contract-price revision in next 5 trading days — if a downward revision lands, verdict flips; (b) hyperscaler capex commentary in next 7 days — any "throttling memory" breaks the demand-side leg; (c) strike settlement in tonight's last-ditch mediation — if cancelled, the YTD-runner unwind likely accelerates and the supply-tightening catalyst evaporates; (d) China CXMT mobile-DRAM final negotiation print (TrendForce noted May timing — nothing leaked yet).
  • Dated watches: strike day 0 = 2026-05-21 (Thursday); strike end = 2026-06-07; next IDC/TrendForce contract-price print due mid-June.
  • Per-ticker (R14 tape):
    • MU → $681.54 (2026-05-18), RSI 67, -14.3% 7D, +52% 30D — 30D leader, 7D weakest, no oversold-bid yet
    • SNDK → $1333.01 (2026-05-18), RSI 67, -13.9% 7D, +46% 30D — cohort-symmetric
    • STX → $740.84 (2026-05-18), RSI 70, -11.2% 7D, +37.3% 30D
    • WDC → $458.68 (2026-05-18), RSI 64, -11.1% 7D, +22.6% 30D
    • HY9H.F → €992 (2026-05-19), RSI 63, -4.6% 7D, +43% 30D — only fresh-19 row; less severe, still red
3 events
DateTypeEvent
May 29 ScanMonster Scan Jun 3 InvestigationFollowing the money up Dell's supply chain — where the alpha moved and what's durable vs cyclical bubble Jun 3 InvestigationNearline HDD storage chokepoint — demand corroborated, supply chain mapped, durable vs bubble and coverage decision

No direct external sources are attached to this read.