Margins Hit 58% and the Guide Implies +69%, but Exabytes Grew 22% for a Second Quarter: Downgrading to Hold on the Trigger We Set in August

Results vs. Consensus

Q1 FY27 Scorecard

Published EPS consensus ranged from about $4.04 to $4.11; the scorecard uses the most widely cited $4.11. Against the low end of the range the beat was $0.60, or 14.9%.

Actual vs. Company Guidance

The guide being graded is the one issued on the Q4 FY26 call in August. Every component landed at or better than the favorable end of its range.

Year-over-Year Comparison

One calendar fact affects every line in that table and was not mentioned on the call. Fiscal 2026 was a 53-week year, and the extra week fell in Q1 FY26, which ran 14 weeks against 13 for this quarter. A straight-line adjustment would lift revenue growth to about 67% and exabyte growth to about 31%. We do not treat the adjusted figures as the real rate. The shipment pattern around the 14-week quarter suggests the extra week added much less than a full week of volume: exabytes rose 7.4% sequentially into Q1 FY26 with the extra week and still rose 5.4% in Q2 FY26 without it. Management itself described the current rate as "slightly below the 25%". The reported 22% understates the like-for-like rate by some amount, and that amount is most likely small. Q2 FY27 compares 13 weeks with 13, which makes it the first clean test of the reacceleration claim.

Sequential Comparison

Management characterized the sequential exabyte gain as 8%; on the rounded unit figures it is 7.4%. Either way, revenue per exabyte rose faster than exabytes for a second consecutive quarter. In Q4 FY26 revenue grew 12.3% sequentially on 4.1% more bits. In Q1 FY27 it grew 16.9% on 7.4% more bits. Price did slightly more of the sequential work than volume this time, against roughly two-thirds last quarter. That is an improvement in mix, though not the change the thesis needed.

Revenue: a record for the HDD-only company, with unchanged volume growth

Revenue of $4,381M was 55% above the year-ago $2,818M and 17% above Q4's $3,747M. The 10-Q decomposes the Cloud line, which is 88% of revenue: exabytes up "more than 20%" and average selling price per exabyte up "approximately 25%". The two smaller lines are even more price-driven. Client revenue rose 119% on roughly 15% more exabytes and roughly 90% higher price per exabyte. Consumer revenue rose 19% even though exabytes fell about 30%, on price per exabyte up about 70%. The company attributes the consumer volume decline to "general softness in demand in these end markets as a result of the increased pricing levels." That is the first explicit evidence in this cycle of price elasticity in WD's own numbers, albeit in its smallest business.

Read with the five-quarter exabyte record, the revenue line says something different from what our August note argued. We described Q4 FY26 as exabyte growth "halving" from 34% to 22%. The fuller series is 23%, 22%, 34%, 22% and 22% for Q1 FY26 through Q1 FY27. The 34% was the anomaly, flattered by a soft year-ago quarter, and 22-23% is the rate WD has delivered in four of the last five quarters. That makes the bull case harder, not easier. Management's reacceleration claim is not a return to a recent normal. It is a step up to a rate above 25% that WD has reached only once in five quarters, while the principal competitor last reported growth of 34%.

Margins: a second straight quarter of beating the ceiling by 200bp or more

Non-GAAP gross margin of 58.1% expanded 1,420bp YoY and 370bp sequentially, clearing the guide ceiling by 210bp after a 240bp beat of the ceiling in Q4. Year-over-year incremental gross margin was 83.6%; sequential incremental gross margin was 80.0% against about 68% implied by the midpoint of the original Q1 guide. Management named three drivers: pricing (contractual LTA step-ups, better spot pricing for nearline customers not on LTAs, and broad strength in client and consumer), mix toward higher-capacity drives, and areal density that delivers more exabytes without the fixed cost of more units.

Non-GAAP operating expenses of $391M were up 3% YoY on revenue up 55%, so operating expenses fell to 8.9% of revenue from 13.5% a year ago. Operating margin of 49.1% is a level the company had not printed before as a pure-play HDD manufacturer. The Q2 guide takes gross margin to 60-61% and operating margin to roughly 52.6% at the midpoint. These margins are what a supply-constrained duopoly with a small third player looks like at the top of a pricing cycle. They are also what a structurally improved industry would look like. The margin line cannot distinguish the two, which is why the thesis has always rested on the volume line.

EPS: clean, and now a single number

Non-GAAP diluted EPS of $4.64 rose 161% YoY and 30% sequentially. The build is $2,153M of non-GAAP operating income, $10M of net interest and other expense, a 17.0% tax rate and 383M diluted shares, for $1,779M of net income. GAAP EPS of $4.63 differs only on the share count, because the non-GAAP count credits the capped calls on the 2028 convertible notes, which offset about 1M shares. The August note predicted that GAAP and non-GAAP earnings would converge once the SanDisk interest was gone, and they have. From here, GAAP earnings are a usable valuation input, subject to stock compensation ($79M this quarter) and business realignment charges ($43M, with about $74M more expected in the rest of FY27).

The share count deserves a note. The convertible notes were largely settled in the quarter: $191M of principal was exchanged for cash plus 4.7M shares, and $378M was converted with the premium settled in 9.3M shares. WD repurchased 2.0M shares for $1.00B, at an average of roughly $500. The non-GAAP diluted count fell to 383M from 388M, and Q2 is guided to about 380M. The remaining $141M of notes are called for redemption on November 16.

Segment Performance

WD reports three end markets. The 10-Q now discloses Cloud revenue to the million, so the table needs no residual.

Year-over-year growth rates compare a 13-week quarter with a 14-week quarter. The Q4 FY26 Cloud comparison base of $3,335M is derived from total revenue less the disclosed Client and Consumer figures.

Cloud

Cloud revenue of $3,868M grew 54% YoY and 16% sequentially, slipping to 88% of revenue from 89% as the Client line outgrew it. The 10-Q split, more than 20% from volume and about 25% from price, shows that even inside the contracted nearline business price contributed slightly more than volume. Management described three pricing mechanisms at work: contractual step-ups as LTAs reach their anniversaries, repricing as LTAs expire and are replaced, and spot pricing for nearline exabytes that are not under contract.

"Cloud revenue was $3.9 billion, representing 88% of total revenue, and increasing 54% year-over-year. Driven by durable demand for our high capacity nearline HDDs. Including the initial ramp of our latest EPMR platform."
— Kris Sennesael, CFO

Assessment: The Cloud line is where the thesis is decided, and the 10-Q's decomposition is the most useful disclosure of the quarter. Volume "more than 20%" is consistent with the company-wide 22% and confirms that the shortfall against the 25%-plus demand rate is in the core business, not an artifact of the small lines. The spot-pricing comment matters more than it sounds: it means part of the nearline book now prices at market, which raises both the upside while supply is tight and the speed of any downturn when it is not.

Client

Client revenue of $320M more than doubled YoY and grew 42% sequentially, on about 15% more exabytes and about 90% higher price per exabyte. In Q4 management tied this line's strength to higher pricing on competing flash products. This quarter the 10-Q cites mix toward higher-capacity drives and an improved pricing environment, and the flash shortage is the likely force behind the latter.

Assessment: This is shortage revenue and should be modelled as such. At 7% of revenue it added about 6 points to total YoY growth this quarter. It also carries the cleanest signal of where pricing sits in the cycle: a 90% increase in price per exabyte is not a contractual escalator, it is a scarcity premium, and it is likely to be the first line to reverse when flash supply loosens.

Consumer

Consumer revenue of $193M grew 19% YoY but only 3% sequentially, as exabytes shipped fell about 30% YoY against price per exabyte up about 70%. The 10-Q explicitly attributes the volume decline to softer demand caused by the higher prices.

Assessment: At 5% of revenue the dollars are immaterial. The read-across is not. This is the one end market where buyers have no LTA, no qualification lock-in and no supply-assurance motive, and it is showing that a 70% price increase destroys about 30% of volume. Nearline buyers are far less price-elastic, but the line is a reminder that the current price trajectory has a ceiling somewhere.

Key KPIs

The inventory build is worth decomposing. Of the $270M increase, raw materials and components rose $150M (to $435M from $285M) and work-in-process $106M, while finished goods rose only $14M. That is not a channel or finished-goods build. It is consistent with WD securing components in a tight supply chain, including the DRAM and NAND whose costs management flagged, and with longer HAMR production cycle times. Receivables of $2,232M rose 10% against 17% revenue growth, so days outstanding improved to 46 from 49.

Key Topics & Management Commentary

Overall Management Tone: Management was more specific on milestones than in August, giving dated qualification targets, interim adoption percentages for the 40TB platform and an explicit statement that this quarter marks the floor for exabyte growth. On the two soft spots, volume and cost per terabyte, the posture was explanation rather than quantification, and on competitors' capacity plans it was argument from vertical integration rather than data. Tone was more confident than in August on demand and pricing, and no more convincing on the volume line.

1. Exabytes grew 22% again, and management called it the floor

WD shipped 248 exabytes, up 22% YoY and 7.4% sequentially on the rounded figures. Management attributed the rate to timing between two product transitions: WD executed the up-to-32TB ePMR transition over the last couple of quarters, and the 40TB ePMR and HAMR transitions are just beginning. The commitment was directional and strong: this quarter is the floor, growth accelerates through FY27 and into FY28, and, in answer to a direct question, roughly 35% exabyte growth within about six quarters is "the right direction."

"And, again, we have indicated that currently, it is slightly below the 25% because we are in between those 2 product transitions. But as the product transitions are getting more traction, we will exceed that 25%."
— Kris Sennesael, CFO

The August note set two markers for this print: exabyte growth back above 25% would restore High conviction, and a second quarter in the low twenties with price carrying revenue would move the rating to Hold. The second one happened. The 14-week comparison adds some unknown upward adjustment to the like-for-like rate, but management's own characterization treats 22% as the real number.

Assessment: The explanation is the same one offered in August, now with better milestones attached. The transition gap is real and mechanically plausible. The problem is that the explanation now has to cover a second quarter, and that the competitor's higher growth rate shows the industry can ship more bits than WD is shipping. This is the single reason for the downgrade.

2. Price per terabyte accelerated to +28%, and Q2 implies more

Blended average price per terabyte rose 28% YoY, up from the high teens (18%) in Q4 and high single digits in Q3 FY26. The Q2 guide of $5.1B is 69% above the year-ago quarter. Unless exabyte growth jumps to 30% or more, that requires revenue per exabyte to grow in the mid-to-high 30s: about 38.5% if exabytes grow 22%, about 35% if they grow 25%. The CFO confirmed that price-per-terabyte growth has further room "for the next 3, 4 quarters" and expects it to slow "eventually."

"Year over year growth in average price per terabyte further increased to 28% versus the high teens we reported in the prior quarter."
— Kris Sennesael, CFO

"And in addition to that, of course, some of our exabytes are not under contract as well, and we do see a very strong price environment for those exabytes that are not under contract as well."
— Kris Sennesael, CFO

Assessment: Pricing is moving faster than any contract-escalator framework would produce, and management now openly describes part of the nearline book as priced at spot. That is excellent for the next three or four quarters and is the clearest signal yet that this is a shortage-pricing cycle. It also hands the bear case a date. The CFO's own expectation is that price growth peaks in three to four quarters, at or near the end of this rating's 12-month horizon. Memory and storage stocks have tended to peak around the turn in the rate of price increases rather than the turn in price itself.

3. Gross margin of 58.1% and a guide to 60-61%

The margin beat came from the same three sources as in Q4, with pricing larger. Management cited the mix shift to higher-capacity drives, operating discipline and a favorable pricing environment across the portfolio. The sequential incremental gross margin of 80% beat the roughly 68% the original guide implied at its midpoint, and the Q2 guide implies about 75%.

"Gross margin reflects a mix shift towards higher capacity drives, disciplined operational execution, and a favorable pricing environment across our portfolio."
— Kris Sennesael, CFO

Assessment: Bull 2 is not in question. WD has now beaten its gross-margin guidance ceiling by more than 200bp in two consecutive quarters, and the Q2 guide adds another 240bp at the midpoint. The relevant question is no longer whether margins expand but what share of the expansion is price, which would reverse, and what share is mix and density, which would not. Management did not split it. With cost per terabyte falling only 4-5%, the arithmetic says most of this quarter's expansion was price. As a rough check, price up 28% against cost down about 4.5% takes the year-ago cost ratio of 56.1% to about 41.9%, which is the 58.1% margin almost exactly. On that approximation price alone accounts for roughly 1,230bp of the 1,420bp gain.

4. Cost per terabyte fell only 4-5%, on DRAM and NAND inflation

A figure put to management on the call, cost per terabyte down about 4-5% YoY in the September quarter, was not disputed. The CFO attributed the shortfall to input cost increases on the DRAM and NAND that sit inside every hard drive, said value-engineering programs are under way but take time to qualify, and guided that cost per terabyte will decline by less than 10% in the near term.

"And so we are well on track in the long run to reduce our cost per terabyte approximately 10% year-over-year."
— Kris Sennesael, CFO

Assessment: This is the second consecutive quarter of a widening miss against the ~10% target, from 8% to 4-5%. It is invisible in reported margins only because price is rising about six times faster than cost is falling. The spread is the whole margin story: when price growth slows toward the low single digits, a 4-5% cost decline produces a very different gross-margin trajectory from a 10% decline. The input-cost explanation is credible, since the memory shortage is real and affects every drive maker. It also means WD's cost curve is temporarily exposed to the same memory cycle that is likely inflating its client revenue.

5. The 40TB ePMR ramp is ahead of prior transitions

The most encouraging operational disclosure. The 40TB ePMR platform, which entered volume production with two customers in the June quarter, is ramping faster than any previous ePMR transition. Management gave interim targets for the first time: close to 30% of nearline shipments in Q2 and 50% in Q3 FY27.

"Customer adoption is progressing ahead of prior transitions. And we expect this platform to represent close to 30% of our nearline shipments by the second quarter and 50% by the third quarter of fiscal 27."
— Irving Tan, CEO

Assessment: This is the mechanism behind the reacceleration claim and it is now checkable one quarter out rather than two. A platform that moves from two customers to roughly 30% of nearline shipments within two quarters should lift exabytes per unit materially in Q2. If Q2 exabyte growth is still near 22% with the 40TB platform at 30% of shipments, the transition explanation stops working.

6. HAMR: two hyperscale qualifications finishing in CQ1, priced at parity per terabyte

Management restored a quantified HAMR pipeline after dropping it in August. Qualification of the first-generation 44TB HAMR drive is on track to finish with two hyperscale customers in the first calendar quarter of 2027, volume shipments begin in the first half of calendar 2027, two more customers start qualification in the first half, and qualification of a second-generation drive at 5TB per platter on an 11-disk platform (55TB) starts in the second half of calendar 2027. Customer feedback was described as very positive, particularly on failure rates relative to other suppliers. Separately, management said it will not discount HAMR against ePMR: pricing is per terabyte regardless of recording technology.

"On our first generation HAMR product, we are on track to finish qualifications with 2 hyperscale customers in the first quarter of calendar year 27. And are on schedule to begin volume shipments in the first half of calendar year 27."
— Irving Tan, CEO

Assessment: The roadmap is the strongest part of the story and this quarter made it more specific. The cost of that roadmap is time: WD's HAMR volume begins in the first half of calendar 2027, when the principal competitor is on a stated path from about 40% of its nearline exabytes on HAMR in mid-2026 to about 70% by mid-2027. The bit-share gap is likely to persist through at least mid-calendar 2027. Pricing at parity per terabyte is the right commercial choice and removes one bear concern, that early HAMR customers would be bought with discounts.

7. Competitors' capacity plans, answered with vertical integration

October brought two supply-side shocks to the group: a reported plan by the smallest of the three drive makers to double its HDD capacity and lift its capacity share toward 30%, and reports that rivals were bidding for the only independent maker of recording heads, which supplies all three drive makers. Management declined to comment on others' plans and argued that bit growth is constrained by heads and media, not by units, and that a maker without its own head and media technology must add units to keep share.

"Other organizations that do not have that degree of vertical integration I would say, will have a harder time being able to reach the same levels of growth that we are seeing."
— Irving Tan, CEO

Assessment: The argument is technically sound. Exabyte supply is set by areal density, and doubling unit capacity at a lower capacity point adds fewer bits than the headline implies. It is also the argument every incumbent makes when a competitor adds capacity, and it does not address the pricing effect of more bits arriving in a market where part of the book now prices at spot. The answer on head sourcing was reassuring for the high-capacity lines (all on internal heads) and left open how much of WD's lower-capacity output depends on the external supplier now in play.

8. Capex steps above the 4-6% range for two years

Capital spending of $227M was 5.2% of revenue, up from 2.9% in Q4. Management set a five-year average of 4-6% of revenue, with the next two years modestly above that range, directed at head and media technology and factory automation and explicitly not at unit capacity. HAMR heads require more process steps and longer cycle times, and the roadmap to drives with up to 14 platters requires more heads per drive.

"While we expect capital expenditures to average 4% to 6% of revenue, over a 5 year period. We anticipate that CapEx as a percentage of revenue will be modestly above that range over the next 2 years."
— Kris Sennesael, CFO

Assessment: On our FY27 revenue estimate, "modestly above" 6% means roughly $1.3B or more of capex against about $0.4B annualized in Q4 FY26. That is a meaningful step down in free-cash-flow conversion: at Q1's 36.6% operating cash flow margin, capex of 6.5% of revenue leaves an FCF margin near 30%, against 34.2% in Q4. The money is going to bits rather than units, which management correctly distinguishes. But head and media capacity is bit capacity, and the industry's supply discipline argument rests on nobody adding bits faster than demand. A question on the call made this point directly, and the answer was that more heads are needed per drive rather than more drives.

9. Visibility to 2030 across the top seven customers, and a broader base

Management said it has good visibility to 2030 across its top seven customers and that the demand signal has risen quarter after quarter. Demand is broadening beyond hyperscale into enterprise, neoclouds, frontier AI labs and hybrid-cloud OEMs. The 10-Q quantifies the broadening: the top ten customers were 64% of revenue against 73% a year ago, on a revised end-customer basis applied to both periods, and two customers exceeded 10% of revenue against four a year ago.

"In fact, we see quarter after quarter, the demand signal actually increasing over time."
— Irving Tan, CEO

Assessment: The concentration data is the most underappreciated disclosure of the quarter and the one clear improvement in the risk profile. A nine-point drop in top-ten share in a year, during a supply shortage when the largest buyers have the most allocation leverage, means WD is genuinely selling to a wider market. It does not change the volume question, but it lowers the probability that one hyperscaler's purchasing pause becomes WD's quarter.

10. Four demand drivers, and the neocloud question

Asked about the risk of a neocloud funding slowdown, management laid out four demand drivers: secular cloud growth (now accelerating on video and AI-enabled services), model training and reinforcement learning, inference and agentic AI (logs, checkpoints, model outputs), and physical AI. The argument is that softness in any one is absorbed by the other three, and that WD selects neocloud customers for balance-sheet strength.

"But we do not see that being in any way a major hiccup if there is a bit of a slowdown in the neo-cloud segment."
— Irving Tan, CEO

Assessment: The diversification argument is reasonable and consistent with the concentration data. It is also an answer to a question the market is asking more often: whether AI-infrastructure financing is the marginal source of storage demand. Management's framing places inference and agentic workloads, not training, as the growth engine, which is the correct frame for a long-duration storage thesis and the harder one to verify.

11. High-bandwidth drives: still nascent

The dual-pivot high-bandwidth drive is sampling with multiple customers, with positive feedback and identified use cases including AI inference caching and video streaming. Management called the opportunity nascent today and a differentiator in the latter part of calendar 2027 and beyond, and said the commercial and pricing construct is still being worked out.

"Look, I think the opportunity is still somewhat nascent today."
— Irving Tan, CEO

Assessment: Unchanged from August: real optionality, no revenue timeline, outside the model. The customer count was not updated from five.

12. Capital return resumes its growth, funded partly by the revolver

The board raised the quarterly dividend 17% to $0.175. Since the program began at $0.10, the dividend has risen every other quarter ($0.125, then $0.15, now $0.175), so the August hold was the pattern rather than a pause. Our August note described it as following three consecutive raises, which was wrong. WD repurchased 2.0M shares for $1.00B and paid $56M of dividends, a total return of about $1.06B against free cash flow of $1,377M. It also spent $570M settling convertible notes and $317M on taxes for vested employee stock. To fund all of it, WD drew a net $300M on its revolving credit facility, which now carries $650M and matures in January 2027. Net cash was flat at $525M.

"Supported by the strength of the free cash flow generation, and consistent with our capital allocation framework, The board has approved a 17% increase in the quarterly cash dividend."
— Kris Sennesael, CFO

Assessment: The dividend raise resolves one of the August watch items, and on the every-other-quarter pattern it was the expected step. The buyback is less impressive in hindsight: $1.00B at an average near $500 is about 45% above the current price, bought in the months before the October slide. That is a timing observation rather than a governance concern. Total financing uses of about $1.94B (buybacks, dividends, convertible settlements and employee tax payments) exceeded the quarter's $1,377M of free cash flow, and the gap came from cash and the revolver. The $650M revolver balance due in January is small against $1.3B of cash and $1.4B of quarterly FCF, and it is the main near-term maturity on the balance sheet.

Guidance & Outlook

The guide is internally consistent. $5,100M of revenue at a 60.5% gross margin is $3,085.5M of gross profit. Less $405M of operating expenses that is $2,680.5M of operating income, and less $15M of interest, taxed at 17%, it leaves $2,212M of net income, or $5.82 on 380M shares. The guided $5.80 matches with no unexplained residual. The opex increase is attributed to qualification activity, consistent with the HAMR schedule.

The guide is also extraordinary in scale. Revenue growth of 69% YoY on a like-for-like 13-week comparison accelerates from 55% reported in Q1. Gross margin rises another 240bp at the midpoint. The YoY incremental gross margin implied is about 81%, and the sequential incremental about 75%.

Street at: pre-print Q2 consensus was about $4.57B of revenue and $4.82-4.84 of EPS. The guide midpoint is 11.6% above on revenue and about 20% above on EPS. For reference, our own August build had Q2 at $4.35B and $4.36.

What the guide requires: revenue per exabyte up roughly 35-39% YoY unless exabytes accelerate well beyond 25%. The guide is a pricing guide first and a volume guide second, which is the composition our thesis was built to avoid depending on.

Guidance style: conservative, and now demonstrably so for five quarters. WD has landed at or above the top of its revenue range in each quarter since Q1 FY26 and beaten its gross-margin ceiling by 200bp or more twice in a row. On that record Q2 is more likely to land near $5.2B and 61% than at the midpoint. Guidance credibility is not the issue. Durability is.

Analyst Q&A Highlights

What stands between HAMR qualification and volume, and what sets the ramp speed

The call opened on HAMR rather than on the quarter, asking for the milestones between finishing qualification in March and volume shipments by mid-2027, and whether the ramp is paced by customer deployment or by WD's own manufacturing. Management answered on confidence, citing customer feedback on failure rates relative to other suppliers, and on supply-chain readiness. It did not name a gating factor.

Q: "Can you just talk about what are the remaining milestones between those 2 steps from your perspective? And then what determines how quickly you scale from there? ... Driven by customer deployment schedules? Or your manufacturing readiness?"
— Amit Daryanani, Evercore

A: "As we have highlighted in the past, feedback from our customers have been very positive, particularly around the failure rates that we have seen in our drive performance vis a vis what they have seen with other players. ... As a result of that, we are very confident in terms of finishing qualifications ... We are working closely with our own internal manufacturing assets and vendors to get the supply chain ready to support the ramp as well."
— Irving Tan, CEO

Assessment: The failure-rate comparison is the most competitively pointed thing management said all call, and it is a direct claim that WD's first HAMR product is more reliable than the incumbent's. The question of what paces the ramp went unanswered, which matters because the bit-share gap closes only as fast as HAMR volume grows.

Why cost per terabyte fell only 4-5%

The follow-up put a number to the cost line, about 4-5% down YoY, and asked how that squares with very strong gross-margin guidance. The CFO accepted the premise, attributed it to DRAM and NAND input costs, and moved the ~10% target to "the long run."

Q: "But I think cost per terabyte was down like 4% to 5% in September quarter. ... You can just touch on what is happening there and what is the right way to think about this longer term because do realize you are guiding for some really strong gross margins in December."
— Amit Daryanani, Evercore

A: "In the midterm, and I have indicated this before, this is not a surprise, we are facing some input cost increases As most of you know, we use some DRAM and some NAND, inside our hard disk drives. And the whole supply chain really operates in a very tight environment. ... It takes a little bit of time to implement and qualify those solutions. And so in the near term, our cost per terabyte will be less than 10% down year-over-year."
— Kris Sennesael, CFO

Assessment: A straight answer, and an unflattering one. The ~10% cost decline was the variable that made the long-run margin case independent of pricing. It is now explicitly a long-run target, with no date on the value-engineering offsets. Modelling should assume the spread between price and cost narrows from both sides within the next year.

Whether a smaller rival's capacity plans change the industry or the LTA negotiations

The question tied the October supply news directly to pricing: whether the smallest player's capacity additions and the consolidation talk affect the industry, and whether they are showing up in out-year price negotiations. Management said fundamentals are unchanged and redirected to its vertical-integration argument. The second half of the question, about pricing in negotiations, was not addressed.

Q: "Irving, I was wondering if you could just address this angst around incremental capacity ads from the smallest player in the market, and there is been some chatter around potential consolidation in the industry. ... And is that impacting any of your price negotiations with hyperscale customers in the out years?"
— Wamsi Mohan, Bank of America

A: "I would say at a high level, we do not see any change in terms of the long term fundamentals of both the industry or our business specifically. ... You know, there may be some other players in the market who do not have that unique ability to build their own media, build out their own heads, drive that area density improvement, and they have to compensate that by potentially maybe increasing the number of units just to maintain their share."
— Irving Tan, CEO

Assessment: The most important question on the call, and the specific part that matters went unanswered. Whether customers are using new supply announcements to push back on 2028-2030 pricing is exactly what investors need to know, and management answered a different question about long-term fundamentals. The silence is not evidence that negotiations have changed. It is a missed chance to say they have not.

Whether the September quarter is the floor for exabyte growth

The follow-up asked whether FY27 as a whole lands near the long-term exabyte range, and how far unconstrained demand exceeds supply. Management gave the clearest forward statement of the call: this quarter is the floor, growth accelerates through FY27 and into FY28, and the supply-demand gap is not closing in the near term.

Q: "So on the exabyte shipments, you grew 22%. ... Like, is it right to think of fiscal 27 at roughly your longer term range for the year?"
— Wamsi Mohan, Bank of America

A: "We sort of see the current quarter as being sort of the floor in terms of exabyte growth, and we see that accelerating through the rest of the fiscal year. ... There has not been any diminishing in demand, and we do not see the gap in supply or demand that exist today is perishable in the near term as well."
— Irving Tan, CEO

Assessment: "The floor" is a falsifiable statement with a one-quarter horizon, and that is its value. It converts Q2 into a test: exabyte growth at or below 22% would contradict management's own floor, and growth above 25% would vindicate the transition explanation. It is also the statement we would most like to believe and are least willing to price in advance, having heard the same direction in August.

Closing the exabyte gap with the principal competitor

A direct comparison: WD in the low-to-mid twenties against the peer in the mid-thirties, and whether the goal is to close a ten-point gap. The follow-up asked whether the message is about 35% exabyte growth within six quarters. The CFO repeated the one-quarter guidance policy and then said the questioner was thinking about it "in the right direction."

Q: "You have been undershipping your peer by about 10 points. You are sort of mid twenties, and they are sort of mid thirties. ... So, Christopher, should we think of you in 6 quarters shipping, like, 35% exabyte growth? Is that the message?"
— Tim Arcuri, UBS

A: "So we only guide 1 quarter at a time, but you are thinking about it in the in the right direction."
— Kris Sennesael, CFO

Assessment: An implicit acknowledgement of the gap and a soft endorsement of a 35% target about six quarters out, which would be around the March 2028 quarter. That is a long time to carry a bit-share deficit in a cycle whose price growth, by the CFO's own estimate, peaks within three or four quarters. The timing mismatch, with price growth fading before volume growth arrives, is the core of the Hold case.

How long price-per-terabyte growth can run above the high teens

The question noted that WD had historically described high-teens price growth as sustainable and has now run well above it for two quarters, and asked about sustainability. The CFO gave the most specific forward pricing statement WD has offered: further room to increase the rate of growth for three to four quarters, followed by an eventual slowdown.

Q: "So you guys have historically kind of said, high teens is the sustainable growth rate. You are now well above that for a couple of quarters. Do you think you could give us an update about what the right way to look at that is in terms of sustainability of that pricing increases on a go forward basis?"
— Tom O'Malley, Barclays

A: "Just in the 1 quarter ago, we were already at 18% ASP increase on a ASP per terabyte basis. And so last quarter, we did 28. ... I would assume eventually that is going to slow down a little bit. But for the next 3, 4 quarters, there is definitely further room to expand our, ASP per terabyte year-over-year growth."
— Kris Sennesael, CFO

Assessment: This answer dates the pricing cycle. Three to four quarters of further acceleration puts the peak in price growth around mid-calendar 2027, at or near the end of this rating's 12-month horizon, while HAMR volume is only beginning. For a stock that trades on the second derivative of price, the most bullish pricing commentary on the call also defines when the tailwind turns.

Whether adding head capacity is adding capacity by another name

A deliberate devil's-advocate challenge to the "no unit capacity" framing: more heads mean more capacity per unit, so is WD adding bit capacity even if it is not adding units? Management's answer explained the need rather than disputing the premise: HAMR heads take more cycles to make, and the roadmap to drives with up to 14 platters needs more heads per drive.

Q: "I am just gonna try to play the devil's advocate. By adding head capacity, you are, you know, increasing the capacity. ... So I am just wondering does not that lead to the fact that you are actually adding capacity even though if it is not a unit, leading to more capacity increase."
— Kris Sankar, TD Cowen

A: "And part of it is, as I mentioned, for example, a HAMR head just requires more cycles to produce, so there is a bit of a throughput loss, so we need to offset. ... And we have also indicated at our innovation day back in February of this year we have a road map to deliver drives with up to 14 platters."
— Irving Tan, CEO

Assessment: The honest reading is that WD is adding bit capacity through heads and media, as every drive maker must to grow exabytes, and that the "no unit capacity" line describes the method rather than the outcome. That is fine. Bit growth is the point. But it weakens the claim that this cycle's supply discipline is a physical constraint rather than a choice, because every maker's bit supply is now growing through the same lever.

How much WD depends on external head suppliers

With the only independent head maker reportedly drawing bids from WD's rivals, the question was whether WD uses external heads at all and whether its head investment is meant to remove that dependence. Management confirmed it buys some external heads and media under a "build and buy" approach for benchmarking, used typically in lower-capacity drives, with all high-capacity leading-edge platforms on internal heads.

Q: "I was wondering if you could maybe just clarify with respect to your head strategy currently, whether you use external vendors for any of your products or for surge capacity?"
— Jim Schneider, Goldman Sachs

A: "So we do use some external head capacity that we always have done. We do that for media as well. ... Going forward, what we do buy from third parties are typically used in our lower capacity drives. So all the high cap leading edge platforms that we introduce are all on our own internal heads."
— Irving Tan, CEO

Assessment: Reassuring where it matters. The nearline platforms that carry the margin are on internal heads, so a change of ownership at the external supplier would not touch the core. The share of total output on external heads was not given, and the client and consumer lines that use them are the ones growing fastest in dollars this quarter.

What They're NOT Saying

1. Any update on UltraSMR. In August management set a target of about 60% of nearline exabytes on UltraSMR exiting FY27, and blamed customer mix between conventional recording and UltraSMR for the volume shortfall. UltraSMR was not mentioned once on this call, and the 10-Q does not mention it either. The mix explanation for low volume was dropped along with the metric that would test it.
2. The 14-week comparison. The year-ago quarter had an extra week, which depresses every YoY growth rate in this print, including the 22% exabyte figure. Management did not raise it, and no questioner did. That WD did not use an available, favorable adjustment is a small credibility point in its favor. It also means the reported 22% is all anyone will grade.
3. Whether new supply announcements are affecting out-year pricing. Asked directly whether a rival's capacity plans were affecting price negotiations with hyperscalers in the out years, management answered on long-term fundamentals and vertical integration and did not address negotiations.
4. How much of the pricing is contracted. Management said that LTAs carry pricing components, that some volume is upside rather than base, and that coverage thins further out. It gave no share of FY28 exabytes under LTA pricing. The CFO's comment that uncontracted exabytes see "a very strong price environment" implies a meaningful spot exposure that is not quantified.
5. The size and duration of the input-cost headwind. DRAM and NAND inflation was cited as the reason cost per terabyte fell only 4-5%, with value engineering as the offset. Neither the dollar impact nor a timeline for the offset was given.
6. Consumer volume down about 30%. The call described consumer revenue as up 19% and cited pricing strength in client and consumer. The 10-Q states that consumer exabytes fell about 30% on demand softness caused by higher prices. The first evidence of price elasticity in WD's numbers appeared only in the filing.
7. The revolver. WD drew a net $300M on its revolving credit facility, which now carries $650M and matures in January 2027. Neither the borrowing nor the refinancing plan was discussed on the call; both are in the 10-Q.
8. Realignment charges at record margins. The quarter carried $43M of business realignment charges, all of it employee termination benefits, with about $74M more expected in the rest of FY27, per the 10-Q. The call did not mention the program. The amounts are small, but restructuring headcount at a 49% operating margin is a choice worth explaining.

Market Reaction

• Pre-print setup: WD closed at $354.19 on October 22, the day of the release, after falling 6.4% in that session from $378.51 as the Nasdaq fell 2.6% and memory and storage names fell 6-9%. The stock entered the print up 105.6% year to date and 194.0% over twelve months, but down 23.8% over the trailing 30 days from $464.60 and 52.5% below its 52-week closing high of $746.23. The October slide began on October 2, when the shares fell 10.2% to $415.29 from $462.56 on the reported capacity expansion by the smallest drive maker, and continued on October 6 (-6.9% to $411.04) on the reported bidding for the independent head maker.
• After hours: the shares initially rose as much as roughly 4% on the release, then reversed during the call and traded about 5% lower, near $337.01.
• Reaction session (October 23): opened at $346.10, a 2.3% gap down, traded between $339.55 and $360.66, and closed at $343.67, down 3.0% or $10.52. Volume was 14.0M shares against an 8.6M 30-day average, 1.6 times normal.
• Index and peers: the S&P 500 fell 0.5% on the reaction session. The principal HDD competitor, which reports on October 27, fell 1.8%, as did the leading flash maker; the largest US memory maker rose 2.5%.
• Since our August note: from the $434.30 close on which the Outperform was reiterated, WD is down 20.9%. Over the same period the principal HDD competitor is down 22.4% and the S&P 500 1.7%.

The reaction to the print itself was modest. A 3.0% decline on a quarter that beat every line and guided EPS 20% above the Street is a weak response, but it followed a 23% slide in three weeks, and the intraday high of $360.66 shows buyers tested the print before sellers won. The after-hours reversal during the call, rather than on the release, points at what was said: a second quarter of 22% exabytes, cost per terabyte at 4-5%, capex above its range, and no answer on whether new supply is changing pricing talks.

The larger move is the one that preceded the print, and it is a group move, not a WD move. Since our August note, WD and its principal competitor have fallen by almost the same amount, about 21-22%, while the S&P 500 barely moved. In October alone the HDD and memory names fell 22-32%. The market has stopped pricing these companies on the quarter and started pricing the cycle: when does price growth peak, and what does new supply do to it. On that question this print offered no relief, because the strongest numbers in it were price numbers.

We have to be direct about our own record. In August we called the 13% post-print decline "a multiple event, and multiple events are where the opportunity usually is." The multiple kept compressing. On the FY27 estimate we now use, the August price was 18.1x and today's is 14.3x, so the de-rating continued by about another fifth even as the estimate rose by about a third. The Outperform cost 19.2 points against the S&P 500 in eleven weeks. That does not by itself argue for a downgrade, since a cheaper stock with higher earnings can be a better buy. The downgrade rests on the volume condition we set, not on the price.

Street Perspective

Debate: Is 22% exabyte growth a transition gap or lost share?

Bull view: WD is between two platform transitions, and the next one is arriving faster than any before it. The 40TB ePMR drive goes to roughly 30% of nearline shipments in Q2 and 50% in Q3, HAMR qualifications finish in the first calendar quarter, and management has called this quarter the floor. Demand is not the constraint: price per terabyte rising 28% proves customers want more bits than the industry can ship.

Bear view: Two straight quarters at 22% against a competitor in the mid-30s is lost share, whatever the reason. WD's HAMR volume arrives as the competitor moves toward about 70% of its nearline exabytes on HAMR, so the gap persists into 2027. The mix explanation from August has quietly disappeared, and the reacceleration remains a promise for a third quarter.

Our take: The bull case has the better mechanism and the bear case has the better evidence. The 40TB interim targets make Q2 a clean test, on a 13-week comparison, of management's own floor claim. We would rather pay for that evidence after it arrives than before.

Debate: Can price per terabyte keep accelerating as new supply is announced?

Bull view: Supply is set by areal density, and a maker without its own head and media technology adds units at lower capacity points, which adds fewer bits than the headline implies. LTA discussions extend to 2029-2031, the supply-demand gap is described as not closing, and the CFO expects further acceleration in price growth for three or four quarters. The Q2 guide already embeds it.

Bear view: Pricing is now partly spot, the client and consumer lines are pricing on a flash shortage that will end, and consumer volumes are already falling 30% on price. A third player announcing a doubling of capacity is how every prior HDD pricing cycle began to turn. Even management dates a slowdown within a year.

Our take: Pricing is safe for the next two to three quarters, and the market knows it. What matters for a 12-month view is that, by management's own estimate, the rate of increase peaks at or near the end of that window, while WD's volume recovery arrives late. Storage stocks have rarely rewarded holders through the turn in price growth.

Debate: Is 14x FY27 earnings a value entry or a peak-cycle multiple?

Bull view: At $343.67 the shares trade at about 14x FY27 earnings that could reach $24, and below 15x the Q2 guide annualized, with 49% operating margins rising toward 53%, net cash and a growing dividend. Most sell-side targets sit far above the price even after cuts. A business with these margins and contracted demand to 2030 should not trade at a discount to the market.

Bear view: Fourteen times is a generous multiple for earnings that are mostly shortage pricing. If price growth turns in mid-to-late 2027 while cost per terabyte falls only 5%, FY28 earnings could be flat or lower, and at a peak the right multiple for a hardware cycle is single digits. The stock fell 54% from its high while estimates rose, which is what a market pricing a peak looks like.

Our take: The valuation is the reason this is a Hold and not an Underperform. At 14x, a great deal of the cycle risk is already in the price, and the near-term earnings momentum is real. But the multiple is not obviously wrong for earnings of this composition, and the catalyst that would justify paying more, volume growth above 25%, has failed to appear twice. Neutral is the honest position until it does.

Model Update

Valuation impact: Fair value moves to $336 to $420, from $465 to $575. The estimate rises by about a third, and the multiple we are prepared to pay falls by about half. Both changes have the same cause: the quarter showed that a growing share of WD's earnings is price, which is worth more today and less in a multiple. The low end, 14x, is roughly where the market prices the stock now on our estimate. The high end, 17.5x, is just below the 18.1x the August price represented on the same estimate. We would not pay that again without the volume evidence.

Against the October 23 close of $343.67, the range implies 2.2% downside at the low end and 22.2% upside at the high end, with the $378 midpoint implying about 10% upside. That is roughly a market return with above-market volatility, which is the definition of a Hold. The shares trade at 14.3x FY27E against 18.1x at the August note's price and 31.1x at the 52-week high, all on the same estimate.

Thesis Scorecard Post-Earnings

August Commitments Revisited

The August note listed eight forward commitments to grade this quarter. Three were met or improved, one is on schedule, two went the wrong way, and two were left unresolved.

Pillar Scorecard

Overall: Execution strengthened again, and durability evidence weakened for a second quarter. Four of five bull pillars held or improved, and one bear point clearly eased. But the pillar that defines the thesis, WD's participation in structural capacity growth through volume, failed its test a second time. The two bear points about pricing and competition both escalated without yet doing damage to reported results. The thesis is not broken. It is now a thesis about the next two quarters rather than the last two.

Action: Downgrading to Hold from Outperform, conviction 5 of 10. The August note committed in writing to this action if exabyte growth stayed in the low twenties with price carrying revenue, and we are honoring that commitment rather than finding a reason to defer it. Valuation stops us short of a more negative view: at 14.3x FY27E with earnings still rising, much of the cycle risk is priced. The upgrade path is specific. Q2 FY27 exabyte growth above 25% YoY on a clean 13-week comparison, with the 40TB platform near 30% of nearline shipments, would restore the Outperform. A Q2 figure at or below management's own 22% floor, or price-per-terabyte growth that stops accelerating before volume does, would move us lower. A price near 12x FY27E, about $290, would make the risk/reward attractive even before the volume evidence arrives.