R31-B — Freshness check + equipment-layer bifurcation + cohort-state reconciliation

Update

R31-B — Freshness check + equipment-layer bifurcation + cohort-state reconciliation

  • Type: refresh (subtype: freshness-check + equipment-layer-bifurcation-cite)
  • event_id: 2026-05-23-optical-supercycle-r31b-freshness-and-bifurcation
  • Source: R31-B authorized perspective-touch session — workflow/RUNS.md 2026-05-23 r31b-r26c-ai-capex-perspective-touch. Source ticket: R26-C TASKS-PERSPECTIVES.md (R26-C through R29-C extensions). Scan substrate: R30-A precomputed summaries{optical-supply-chain,ai-infrastructure}.json` (fresh 2026-05-22 bars); R28-E equipment-layer bifurcation finding. README cohort-state edit (R23-G decision 4) was landed by R31-A same-session above — this entry is the supporting log evidence + reconciliation.
  • Freshness check (R24-C audit pattern; staleness window 14d): last_refreshed was 2026-05-18 → bumping to 2026-05-23 (5 days; well within the freshness window). The perspective is not stale; this is a normal freshness-cadence touch, not a stale-refresh. R31-A's README cohort-state edit ("hardens with edge decay" → "contested / late-stage") is the substantive change; this log entry documents the underlying tape supporting the cohort-state language.
  • Wave 1 cohort (LITE / COHR / CIEN / FN) fresh-5/22 numbers (optical-supply-chain):
    • LITE RSI 53.1 / +7.72% 7d / +11.81% 30d / +40.34% 3m — consensus zone, not extended; positioning split (R20 13F: Aschenbrenner exited $479M / Druck initiated small).
    • COHR RSI 59.6 / +5.35% 7d / +11.81% 30d / +51.7% 3m — same shape; Q3 print fundamentals still hardening but positioning contested.
    • CIEN RSI 60.8 / +10.69% 7d / +13.16% 30d / +69.28% 3m — quietly leading Wave 1 on 3m magnitude.
    • FN RSI 54.9 / +5.17% 7d / +2.09% 30d / +22.05% 3m — Wave 1 laggard; not extended.
    • Cohort-state read: consistent with R31-A README edit "contested / late-stage." Demand-leg fundamentals (COHR Q3 print) still hold; positioning at smart-money flow level remains visibly contested (R20 13F evidence carries). Edge is decayed.
  • Wave 3 cohort (TSEM / ALAB / SMTC / MTSI) fresh-5/22:
    • ALAB RSI 79 / +42.58% 7d / +55.35% 30d / +139.3% 3m — momentum-extended hard; near short-term blow-off shape.
    • TSEM RSI 67.8 / +13.12% 7d / +40.82% 30d / +122.42% 3m — Wave 3 momentum still going; not yet extended on RSI.
    • SMTC RSI 76.8 / +17.06% 7d / +48.83% 30d / +80.58% 3m — extended.
    • MTSI RSI 73.6 / +6.63% 7d / +35.72% 30d / +57.2% 3m — extended.
    • Cohort-state read: Wave 3 is the momentum layer exactly as the README projection has it. ALAB at RSI 79 is the front-edge; partial pullback risk into next earnings cycle is the natural reading. NOT a thesis weakening; cohort-tape-extension is what late-stage rotation looks like.
  • Wave 4 cohort (LWLG + adjacencies) fresh-5/22:
    • LWLG RSI 48.9 / +8.53% 7d / -5.51% 30d / +264.02% 3m — speculative leg; 30d cooled from a parabolic 3m. NOT yet binary-resolved. SilTerra mid-2026 tapeout remains the next material event.
    • SIVEF RSI 69.4 / +38.48% 7d / +137.82% 30d / +372.61% 3m — extended Wave 4 adjacency.
    • AXTI RSI 68.1 / +26.89% 7d / +87.10% 30d / +395.36% 3m — upstream InP supplier; ripped.
    • Cohort-state read: Wave 4 is moving but the binary (LWLG tapeout viability) is still gated to mid-2026; the SIVEF/AXTI move is the upstream picks-and-shovels response, not the modulator validation.
  • Equipment-layer bifurcation (R28-E finding; cross-cohort note): the AI-networking equipment layer that sits one tier above optical-component layer is bifurcating: NOK / EXTR / FFIV / ATEN / ERIC lead +23-104% 3m; incumbent ANET prints +20.87% 3m / -10.73% 30d, materially lagging the equipment cohort it sits in. R29-C nuanced: the lag is 30d/3m, not 7d (ANET participated in the latest weekly leg at +10.25% 7d). Reads as either (A) safe-carry-with-less-to-give-back OR (B) incumbent-disruption-by-challengers. Both readings live; next earnings cycle discriminates. This bifurcation is NOT inside this perspective's cohort (the equipment layer is above optical-component layer), but worth citing for cross-thesis visibility — the same "pricing-in across the vertical" pattern (R31-B ai-power-bottleneck log entry) shows up here too.
  • Aschenbrenner Q2 13F (2026-08-15) watch: the R20 CONTESTED entry flagged that Aschenbrenner exited the entire optical equity cluster ($479M LITE + $89M COHR + $85M TSEM + $21M GLW PUT) without a stated thesis reason. R31-B carries the watch unchanged — Q2 13F will reveal whether he re-establishes any optical exposure (which would partially un-contest) or stays absent (which would harden the contested signal).
  • Effect on thesis: posture unchanged from R31-A README edit (Wave 1 contested / late-stage; Wave 3 momentum; Wave 4 unresolved binary). NOT retired or renamed. The demand-leg fundamentals (COHR Q3 print confirmed) still stand; positioning is split; cohort tape is extended-momentum across Waves 3-4 and consensus-zone in Wave 1. Re-evaluate after the next optical-cohort earnings event (COHR Q4 Aug 2026 is the next material catalyst).
  • No README edit (cohort-state already landed by R31-A this session): R31-A wrote "hardens with edge decay" → "contested / late-stage" in README. R31-B does NOT re-edit the README; this log entry is the supporting evidence layer.
  • last_refreshed bump: README frontmatter + ACTIVE.json both → 2026-05-23 (this row).
5 events
DateTypeEvent
May 23 ArticleWeird-ETF rotation read — MSTR-leverage breakdown, AI-stack narrowing, defensives flat May 25 InvestigationOptical supercycle WHY: what specifically drove the CIEN/COHR demand acceleration May 26 Deep Dive$AEHR (Aehr Test Systems, wafer-level burn-in) May 26 InvestigationCPU Shortage Thesis — First-Principles Derivation May 26 InvestigationSpaceX IPO Deep Dive: Separate Trades + the Oxygen Question

No direct external sources are attached to this read.