{
  "@context": "https://orbyd.app/schemas/dossier.v1.json",
  "ticker": "WDC",
  "name": "Western Digital Corporation",
  "url": "https://orbyd.app/dossiers/WDC/",
  "json_url": "https://orbyd.app/dossiers/WDC.json",
  "status": "HELD",
  "current_conviction": "LOW",
  "graded_conviction": "LOW",
  "archetype": {
    "code": "a2",
    "n": 2
  },
  "current_thesis": "The 29% bounce off the 2026-07-28 low stalled: WDC closed $527.22 on 2026-08-03, −3.23% on a day its sector rose 0.58%, back under a rising 50-day at $562.65. Consensus (~$3.69B / $3.31–$3.35) still sits above the company's own $3.65B ±$100M / $3.25 ±$0.15 guide, so the 2026-08-05 print — 16.8% implied — is the binary.",
  "invalidation_trigger": "A daily close below $462 surrenders the 2026-07-28/29 closing floor ($462.04–$463.51) and re-opens the $422.11 spike low. Secondary: the 2026-08-05 fiscal Q4 print landing at or under the $3.65B / $3.25 guide midpoints, or a September quarter not guided up sequentially against Seagate's ~$4.1B bar.",
  "catalyst_date": "2026-08-13",
  "outcome": "INVALIDATED",
  "outcome_date": "2026-08-06",
  "invalidation_fired": true,
  "themes": [
    "ai-chips-memory",
    "ai-datacenter-infrastructure",
    "semi-foundry-equipment"
  ],
  "tags": [],
  "sources": [],
  "notes": [
    "Fiscal year ends in early July: \"Q4 FY26\" is the June quarter, reported 2026-08-05; \"Q1 FY27\" is the September quarter.",
    "Post-SanDisk separation the revenue base is hard drives. No DRAM or NAND fabs, so Chinese memory headlines hit via cohort correlation, not product.",
    "Kioxia's Yokkaichi NAND joint venture sits with SanDisk, extended through 2034 on 2026-01-29 — WDC-labelled Kioxia headlines predate the separation structure.",
    "Insider sales run through Rule 10b5-1 plans adopted months earlier (director plan 2026-03-05, CEO plan 2025-05-12), so individual Form 4s are scheduled rather than discretionary.",
    "Quarterly dividend is $0.15 per share after a 20% increase announced with Q3 FY26 results on 2026-04-30.",
    "Five-year beta 2.22 on 344.68M shares outstanding; the name moves as a leveraged proxy for the whole storage cohort."
  ],
  "body_markdown": "## Current Thesis\n\nThe print landed above every number the company had promised, and the stock broke anyway. Fiscal Q4 FY26, reported after the close on 2026-08-05: revenue $3.75B, up 44% year over year, against an April guide of $3.65B ±$100M; non-GAAP EPS $3.56 against the company's $3.25 ±$0.15 guide and a Street estimate of $3.29 (Benzinga, 2026-08-06). GAAP gross margin 54.1%, GAAP operating margin 41.7%. The September-quarter guide came in at $4.1B ±$100M revenue (+45% y/y), non-GAAP gross margin 55–56% and non-GAAP EPS $4.00 ±$0.15 — above the ~$4.01B and $3.81 the Street carried.\n\nThe tape's answer: −13.44% on 2026-08-06 (tradingkey market-movers), then a further −3.81% to a $434.30 close on 2026-08-07 (stockanalysis.com). Benzinga's 2026-08-06 write-up put the decline at 19% and called it the worst drop since March 2020; the close-to-close figure for that session was the smaller number. Either way, the $462.04–$463.51 closing floor from 2026-07-28/29 — the level flagged in the prior note as the structure to watch — was surrendered on the print, and the $422.11 spike low of 2026-07-28 is now the nearest thing to a shelf.\n\nWhat an investor is buying here is a contracted hard-drive supply book against AI-datacenter storage demand: CEO Irving Tan told the 2026-08-05 call that customers are asking for long-term agreements running to 2031 and that discussions covering 2029–2031 are already underway on commercial structure rather than volume, with \"visibility remains very strong.\" That leg is intact in the numbers. What is no longer intact is the assumption that beating it gets paid. A beat-and-raise producing a double-digit down day in a name up 488.64% over 52 weeks is an expectations problem, and expectations problems resolve through price, not through the next press release.\n\n## Bull Case\n\n- Q4 FY26 (2026-08-05): revenue $3.75B, +44% y/y, above the $3.65B ±$100M guided on 2026-04-30; non-GAAP EPS $3.56 vs $3.29 Street; GAAP gross margin 54.1%, GAAP operating margin 41.7%.\n- Fiscal 2026 full year: revenue $12.9B, +36%; gross margin +970bps to 49.1%; free cash flow $3.5B (company FY26 results release, 2026-08-05).\n- September-quarter guide of $4.1B ±$100M with non-GAAP gross margin 55–56% marks the first quarter guided above $4B and above the 51–52% margin band guided for June.\n- Tan on the 2026-08-05 call: LTA negotiations extending to 2031, with 2029–2031 conversations focused on pricing construct. That is roughly two years further out than the two-customers-for-2027 / one-for-2028 coverage described in February 2026.\n- Seagate's fiscal Q4 on 2026-07-28: revenue $3.6B (+48% y/y), non-GAAP gross margin 52.7%, September quarter guided to about $4.1B, nearline exabytes allocated into calendar 2028 — the same demand picture from the other side of the duopoly.\n- TD Cowen maintained Buy and raised its target to $540 on 2026-08-06, the only post-print target increase on the tape. The stockanalysis.com average across 26 analysts read $662.13 on 2026-08-09.\n- The 200-day moving average sat at $341.21 on 2026-08-09, well under the $434.30 close, so the multi-quarter uptrend has not been touched.\n\n## Bear Case\n\n- The reaction function has changed. Revenue and EPS beat, the forward guide beat, and the stock lost 13.44% in a session (2026-08-06) with SanDisk −11% and Micron −6% alongside it (24/7 Wall St., 2026-08-06). SK Hynix suffered what Benzinga described the same day as a flash crash.\n- The guide's margin over consensus was thin: a $4.0–4.2B revenue range against a $4.01B estimate. The low end of the range sits below the Street.\n- Post-print sell-side action skewed lower. UBS held Neutral and cut to $525 on 2026-08-06; Rosenblatt held Buy but cut to $800. TD Cowen's raise landed at $540 — roughly a fifth of the $662.13 average target, which now looks like a stack of stale numbers.\n- Price closed 2026-08-07 at $434.30 against a 52-week range of $73.14–$799.87 and a 50-day average of $561.20. Two consecutive closes below the July floor is a structure break, not noise.\n- Short interest of 23.26 million shares, 6.93% of float and 2.24 days to cover (stockanalysis.com, 2026-08-09) is too small to force a covering bid on any bounce.\n- Insider distribution over the trailing 90 days as of 2026-08-04 ran 78 open-market sales across six sellers with zero open-market purchases, including a director selling 3,808 shares on 2026-07-28 at a weighted average of $423.42–$461.89.\n- Sentiment cover is late-cycle: Michael Burry warning on 2026-08-06 that the AI boom could leave \"ghost towns,\" and Benzinga running a \"history offers a bullish signal\" piece on the crash itself the same day.\n\n## Setup & Price Structure\n\nWhere the narrative sits in its life-cycle: **SATURATED**. Dating it — 2026-08-06, when a revenue beat, an EPS beat and an above-consensus September guide produced the largest single-day decline since March 2020. When the fundamental news clears the bar and the bid does not appear, the marginal buyer has already bought. Coverage is fully mainstream (Motley Fool, 24/7 Wall St., Benzinga daily market wraps all carried it), the 52-week gain reads +488.64%, and the sell-side average target of $662.13 sits more than 50% above the market with post-print revisions running down.\n\nThe observable structure as of the 2026-08-07 close of $434.30:\n\n- **Lost:** the $462.04–$463.51 closing floor of 2026-07-28/29. Two closes now sit under it.\n- **Nearest support:** the $422.11 intraday spike low of 2026-07-28. Nothing between it and the current close.\n- **Resistance:** the 50-day at $561.20 on 2026-08-09, which is still above price and will roll over if price stalls here.\n- **Long-term structure:** the 200-day at $341.21, untouched.\n- Market capitalisation $149.70B on 344.68 million shares outstanding; five-year beta 2.22 (stockanalysis.com, 2026-08-09).\n\nCrowding observables, stated without a verdict: retail-facing coverage clustered heavily on 2026-08-06 (at least five distinct Benzinga items on WDC or the storage cohort in one session); price sat far above a rising 200-day and far below the 50-day, a wide-range condition typical of an unwind rather than a base; short interest is low and days-to-cover is 2.24, so there is no mechanical squeeze fuel; insider flow is one-directional under pre-adopted plans. The near-term company catalyst that mattered has passed — the next WDC-specific dated item is a 10-K filing, not a number that resets the story.\n\n## Catalyst Calendar (next 30 days)\n\n- **2026-08-13** — SanDisk Investor Day, 9:00 a.m. ET, with CEO David Goeckeler and CFO Luis Visoso (announced 2026-07-08).\n- **2026-08-14** — Q2 2026 13F filing deadline. The only broad look at which institutions were positioned across the June top and the July drawdown before September.\n- **~2026-08-21 (est.)** — WDC FY26 Form 10-K. Annual customer-concentration disclosure and long-term-agreement language, testable against Tan's 2026-08-05 statement that LTA talks now reach 2031.\n- **2026-08-26** — Nvidia Q2 FY27 results; call 2:00 p.m. PT, quarter ended 2026-07-26 (company announcement). The AI-capex anchor the whole storage cohort has traded against since June.\n- **~2026-08-31 (est.)** — US legislative follow-through on Chinese memory. Lawmakers urged a ban (FT via Benzinga, 2026-07-16) and a Capitol Hill probe followed CXMT's debut (TechTimes, 2026-07-28). No date is fixed and the absence of one is itself the datapoint.\n\n## What Would Change Our Mind\n\nThe structure that mattered has already gone. The $462.04–$463.51 closing floor from late July, named in the prior note as the line, was lost on 2026-08-06 and confirmed lost on 2026-08-07 at $434.30. What is left below is the 2026-07-28 spike low at $422.11, and a **weekly close below $422** surrenders it with no identified shelf between there and the far lower base built earlier in the year.\n\nThe fundamental conditions that would independently break the leg:\n\n- The December-quarter guide, when it comes, not stepping up sequentially from the $4.1B ±$100M September guide — which would mark the September quarter as the peak rather than a waypoint on the 2031 LTA story.\n- Any disclosure in the FY26 10-K that narrows rather than confirms the long-term-agreement coverage described on the 2026-08-05 call.\n- A capex-digestion signal from a top hyperscaler, most immediately readable through Nvidia's 2026-08-26 commentary.\n\nWhat would restore the leg: a weekly close back above the $462–$464 area on expanding participation, which would define the post-print flush as a completed shakeout rather than the first leg of a de-rating. Absent that, the sensible read is that the base has not formed and there is no company-specific event inside 30 days capable of forming one.\n\n## Correlation Notes\n\nWDC trades as one instrument with Seagate, Micron, SanDisk and SK Hynix. On 2026-08-06 the group moved together — WDC −13.44%, SanDisk −11%, Micron −6%, with SK Hynix selling off in Seoul — on two company prints that both beat. Beta of 2.22 means index-level AI-sentiment shocks arrive amplified.\n\nPost-separation the revenue base is hard drives; there are no DRAM or NAND fabs inside WDC, so CXMT's DDR5/LPDDR5 ramp reaches the stock through cohort correlation rather than through product overlap. That distinction was supposed to be arbitraged by the 2026-08-05 print. It was not: the numbers separated WDC from the memory complex and the tape did not.\n\nThe second correlation is to Nvidia as the AI-capex proxy. Nearline demand assumptions are downstream of hyperscaler build plans, which makes 2026-08-26 a larger input to WDC's price than anything WDC itself files this month.",
  "first_seen": "2026-04-19",
  "last_analyzed": "2026-08-09T10:18:32+00:00",
  "last_synthesized": "2026-08-09",
  "last_update_source": "watchlist_research",
  "license": "Content © orbyd. Cite the canonical URL."
}