Western Digital just guided to 69% growth and 60% margins. The stock is down 48%.

The HDD duopoly's pricing power is now doing all the margin work — and the market has stopped paying for it.

WDC · Earnings Call · 2026-10-22

The 69% guide nobody wants

On paper this is the best quarter Western Digital has ever printed. Revenue of $4.4 billion, up 55% year-over-year. Gross margin of 58.1%, up 1,420 basis points. Operating margin of 49.1%. EPS of $4.64, up 161%. Free cash flow of $1.4 billion, of which $1.1 billion went straight back to shareholders, and a 17% dividend raise to $0.175 a quarter. Then the guide: “we expect revenue to be $5.1 billion plus minus $100 million. At the midpoint, this reflects a 69% increase year-over-year” — Kris Sennesael, Chief Financial Officer · 2026-10-22 — with gross margin of 60% to 61% and EPS of $5.80. The equity's response, over three months, has been to fall 48% and sit 52.5% below its 18 June close of $746.23. That is the whole story: a company whose reported results are accelerating into a supply-constrained AI storage boom, and a share price that has decided the cycle is over. Two possible readings. One, the tape is front-running a violent reversion in a two-player industry. Two, the fundamentals are peaking in plain sight.

Margins are now a pricing story, not a cost story

Here is the detail that deserves more attention than the headline guide. Price per terabyte rose 28% year-over-year — up from the high teens the prior quarter — while the cost side of the equation got worse. “in the near-term, our cost per terabyte will be less than 10% down year-over-year” — Kris Sennesael, Chief Financial Officer · 2026-10-22, Kris Sennesael said, versus the ~10% annual decline that has anchored the model for years.

we are facing some input cost increases. As most of, we use some DRAM and some NAND inside our hard disk drives. And the whole supply chain really operates in a very tight environment.
Kris Sennesael, Chief Financial Officer · 2026-10-22

That is a meaningful inversion of the margin formula. A year ago, this company earned gross margin expansion from cost discipline: cost per terabyte was the single strongest keyword on the prior call, and the whole UltraSMR pitch was about a software-driven 20% capacity uplift at no incremental cost. Today that keyword has vanished from the company's top themes entirely, and ASP per terabyte sits near the top of the list instead. The mix has rotated from we engineer the cost down to we charge more for the same bit. Incremental gross margins above 80% on 55% revenue growth are magnificent — and entirely price-dependent. Margins that come from a tight tape are only as durable as the tape.

The truce question, asked four different ways

Analysts circled one anxiety for the whole hour: what happens if someone breaks the supply discipline. Wamsi Mohan asked about “incremental capacity adds from the smallest player in the market” — Wamsi Mohan, Analyst · 2026-10-22; Ben Reitzes asked how anyone gets from 10% to 30% share given the heads and media constraint; Jim Schneider asked, hypothetically, what the response would be. Irving Tan's answer was the same each time, and it is the load-bearing claim of the entire thesis: “we remain committed to not investing in any incremental unit capacity” — Irving Tan, Chief Executive Officer · 2026-10-22. What makes this more than boilerplate is that management is now attaching new money to the old promise. Capex has been running at roughly 3% of revenue — $108 million last quarter against $3.7 billion of sales — and is now guided above the 4% to 6% long-run range for two years, into head capacity and factory automation. Sennesael was emphatic that this is not what it looks like: “in the prepared remarks, we explicitly said that the capex is going to head in media and factory automation and not to expanding our unit capacity.” — Kris Sennesael, Chief Financial Officer · 2026-10-22 The technical logic is sound — HAMR head wafers take more process steps, and a 14-platter roadmap needs more heads per drive. But a rising capital intensity story arriving in the same quarter as reports of a rival funding a capacity expansion is exactly the combination that de-rates a supply-discipline multiple. Note the refrain is unchanged across calls: “we are not adding any unit capacity to our portfolio right now” — Tiang Yew Tan, Chief Executive Officer · 2025-10-30, from a year ago, and “At this juncture, we still do not see any need to increase unit capacity” — Irving Tan, Chief Executive Officer · 2026-04-30, from April. Same sentence, bigger cheque.

What is genuinely new: the visibility bid

The freshest content on this call was not about margins. It was about planning horizons. A quarter ago the company was “in the throes of discussions with customers to establish LTAs for calendar '29, '30 and '31” — Tiang Yew Tan, Chief Executive Officer · 2026-08-05. On Thursday: “we have good visibility up to 2030, 2031 for across all seven” — Irving Tan, Chief Executive Officer · 2026-10-22 of the top customers. That is a step change in contracted duration, and it pairs with diversified customer base language as demand spreads from hyperscalers into enterprises, NeoClouds and AI frontier labs. The operational gap is the other side. Exabytes grew 22% to 248 — below the "greater than 25%" long-run target management keeps reiterating, and well below the mid-30s the peer group is printing. The answer is the product ladder: aerial density via the 40TB ePMR ramp (30% of nearline in the December quarter, 50% by the March quarter) and volume shipment of first-generation HAMR in the first half of calendar 2027, with qualification at two hyperscalers completing in the March quarter and two more starting. Crucially, HAMR is not being bought with discounts: “We are not discounting our Hammer products in any way versus our ePMR products.” — Irving Tan, Chief Executive Officer · 2026-10-22 That is the bull case in one line — a technology transition that adds capacity without resetting price. The balance sheet has already made the turn. Effective net cash of $519 million versus a multi-billion net debt position two years ago, and interest coverage of 231x. Gross profit has doubled off the trough, up 90% year-over-year, while capital expenditure of $108 million stayed modest. The company can afford the head-and-media bill.

The tell is in the tape, not the transcript

Western Digital is not being singled out. The global AI data center complex has rolled over across windows — it sits among the notable decliners on the thirty-day tape with far more negative tickers than positive ones — and a $132 billion hardware name with a 52% drawdown and a 3x price-to-sales multiple was always going to be the highest-beta expression of that de-rate. What the market is charging for is not the 69% quarter. It is the durability of a margin structure that now leans on 28% price per terabyte growth while cost per terabyte decline stalls and the third player builds capacity. Watch two lines next quarter: the spread between ASP per terabyte and cost per terabyte, and whether the capex step-up stays inside heads and media. Everything else is already priced.