Type: refresh (subtype: conviction-gate resolution) · Source: Hoya IR segment disclosure + analyst sources via web (Hoya REPORT 2025 financial highlights; Morningstar Hoya coverage; PortersFiveForce/Hennessy profiles). Massive has no Hoya financials (no results + premium-gated ratios) — operating-company financials sourced from IR/filings per the deep-dive financials rule.

The owed gate: the promote left "HDD-glass share of total Hoya revenue/EBIT owed before HOCPY ranks as conviction vs layer-position inference." User pushed back on deferring it to August ("isn't this a near-term thing?") — correct: the conviction question is structural and answerable from the already-filed Q3 FY26 disclosure; August only adds a marginal refresh, and the un-bid asymmetry is a near-term window. Pulled now.

Data (Hoya FY24, ended Mar 2025, disclosed segment level):

• Life Care ¥550.9B rev (64%), OP ¥95.4B (37% of total OP), 17.3% margin
• Information Technology ¥311.1B rev (36%), OP ¥168.4B (66% of total OP), 54.1% margin
• HDD glass is ONE line inside IT, sharing it with the EUV/ArF mask-blank franchise (~70% global EUV share) + FPD photomasks + imaging. Within-IT product split not publicly disclosed → HDD-glass bounded at ≈11–16% of rev / ~20–29% of EBIT (estimate) — punches above revenue weight because IT runs at 54% margin.

Verdict: HOCPY is diluted and is NOT the conviction nearline pick. (1) HDD-glass is a minority of both rev and EBIT; 64% of Hoya is medical/eyewear. (2) The draft's "~70%" was Hoya's EUV mask-blank market share (+ ~100% overall HDD glass / ~40% nearline-3.5″ glass) — market-share, not Hoya-revenue-from-HDD-glass; the framing implied ~70% glass leverage, which is wrong. (3) Hoya's dominant value driver is the EUV mask-blank near-monopoly, a different thesis (semi-litho / AI-compute chokepoint → optical-supercycle / AI-compute coverage, not nearline). HOCPY's un-bid RSI 48 reflects a JP medical-heavy conglomerate priced on Life Care + the EUV/semi cycle, not a nearline-glass mispricing.

Effect on thesis: HOCPY reclassified in README + key_tickers routing: "best risk/reward upstream pick" → "diluted conglomerate; evaluate on the EUV thesis, not as the nearline pick." Stays in key_tickers (real glass supplier + EUV optionality). The drive makers (STX/WDC, captive media) remain the purest nearline expression — confirms the perspective's own "owning STX/WDC is owning the moat." Financials-waiver: precise HDD-glass sub-split is estimate-only (not disclosed); further precision infeasible from public data and would not change the verdict. ACTIVE.json thesis HOCPY clause updated to match.

Feasibility breadcrumb for future-me: do NOT re-open "pull exact HDD-glass revenue for Hoya" — it is not separately disclosed (segment reporting stops at Life Care vs IT). The bounded ~11–16% rev / ~20–29% EBIT is the answer; the Aug Q4 FY26 print refreshes IT-segment totals but won't break out HDD-glass either.

🔄 2026-06-25 full-scan tape check: CONFIRMING — STX RSI 63 (+21% 30D), WDC 62 (+29%), Toshiba 5991.T strong-up; drive-maker duopoly bid; HOCPY de-rated (RSI 42), consistent with the EUV-thesis-not-nearline reclassification; no capacity-add discipline-break. (Data tables spliced fresh to the 2026-06-25 close; full cross-scan read in 2026-06-25-full-scan-market-brief §6.)

🔄 2026-07-05 operating book: Folded in as a LAYER of memory-supercycle (book PRIMARY 1); status → monitoring. Discipline-break watch counts toward the memory exit tape.