WDC Guided 20% Above Consensus, Yet the Stock Keeps Falling

On the numbers alone, it's hard to find a flaw in Western Digital's quarter. Revenue, gross margin, operating margin and free cash flow all hit their best levels since the company spun off Sandisk, and next quarter's guidance came in well above what the market expected. The key results:

String the last six quarters together and the trend keeps getting steeper:

Gross margin went from 43.9% to 58.1% in a single year. Think of a shop that used to pay $56 in costs for every $100 of goods it sold and now pays only $42. Almost all of that extra $14 drops straight to profit, because non-GAAP operating expenses barely moved over the same stretch, from $381 million to $391 million. That's operating leverage: when sales grow much faster than overhead, profits grow faster still.

One structural detail drives everything that follows. Shipment growth has not actually sped up. It was 34% a year ago and has sat at 22% for the last two quarters. What pushed revenue up 55% was price: 1.22 times the volume multiplied by 1.28 times the price comes out at roughly 1.56 times the revenue. The company gives no full-year guidance, but next quarter's 69% midpoint revenue growth lets us work out what has to happen. If shipments keep growing 22%, price per TB has to rise about 38.5% year over year. Even if shipment growth gets back above 25%, as management expects, price still needs to climb about 35%. Put simply, the guidance assumes price increases step up again from 28%. The single number to watch in the next report is whether price per TB rises at least 35% year over year. That one figure largely decides whether revenue and gross margin hit their targets.

The demand evidence is firmer than it was a quarter ago. On the call, management said demand visibility from its top seven customers now runs out to 2030–2031. The long-term agreements (LTAs) under negotiation run through 2029, 2030 and 2031, and some include contractual price increases. Demand is also spreading beyond a handful of hyperscale cloud providers to enterprise customers, newer cloud providers and frontier AI labs. In the words of CFO Kris Sennesael: "Our first quarter results demonstrate the continued momentum in our business, with robust exabyte growth and improved pricing driving another quarter of strong financial performance." (8-K, 2026-10-22). Management also called this quarter the low point for shipment growth, because the company is moving between two product generations, ePMR and HAMR. ePMR is the current energy-assisted perpendicular magnetic recording technology. HAMR is the next generation, which uses a laser to heat the disk so it can pack data more densely.

Making hard drives doesn't take especially heavy investment. WDC spent $227 million on capital expenditures this quarter, just 5.2% of revenue, and its cash flow covers that many times over. So the real questions here are about the quality of the profit: what separates the GAAP figures from the adjusted ones, and how much of the profit turns into actual cash.

With GAAP and non-GAAP EPS only a cent apart, many readers will assume the profit couldn't be cleaner. For this quarter, that's right. Look back over the past year, though, and the GAAP income statement is far messier. FY2026 GAAP net income was $9.42 billion, a startling figure next to that year's $12.92 billion of revenue. The reason is large gains from selling the Sandisk shares WDC still held: $2.05 billion booked in FY2026Q4 alone, and $611 million in FY2026Q1 (company release). These are one-off investment gains that have nothing to do with selling hard drives, which is why FY2026Q4 GAAP EPS of $8.21 was far above non-GAAP EPS of $3.56. A price-to-earnings ratio built on full-year GAAP figures would badly overstate how much the business really earns, so every valuation in this note uses non-GAAP operating figures.

So did $1 billion of buybacks shrink the share count? No. GAAP diluted shares were 384 million this quarter, up about 2.1% from 376 million a year ago. The share count rose despite the buybacks because WDC's convertible bonds due 2028 moved into the money as the stock soared, and the new shares from conversions more than offset the repurchases. This quarter the company retired a large part of the remaining 2028 convertibles and guided next quarter's diluted share count to about 380 million. It also noted that the guidance doesn't include any extra benefit from the capped call, an options hedge bought alongside the convertibles (8-K and earnings call, 2026-10-22). For existing shareholders, that means about 2% real dilution over the past year, and buybacks only start actually shrinking the share count from next quarter.

A 77% cash conversion rate is decent but not outstanding. The press release doesn't break out tax payments or working-capital swings for the quarter, so we'll need the cash-flow detail in the 10-Q to explain the gap. The drop in quarter-end cash came mainly from buybacks, dividends and convertible repayments all landing at once. Those are deliberate choices about how to use cash, not a sign the business is bleeding money.

WDC is now a pure hard-drive company. Its flash-memory business was spun off as Sandisk in 2025, and that matters for understanding its cost structure.

On the supply side, the two most important parts of a hard drive are the read/write heads and the disk media. WDC largely makes both itself and uses some outside suppliers only for low-capacity products (earnings call, 2026-10-22). The real cost pressure comes from a place that's easy to miss: hard drives also contain DRAM cache and NAND flash chips, and both are getting more expensive. The CFO admitted on the call that the year-over-year decline in cost per TB will be under 10% in the near term, short of the company's long-term target of about 10%. So while WDC enjoys rising prices from its customers, rising memory-chip prices claw a little of that back. The first effect is just much bigger. On capacity, management said capital spending over the next two years will run slightly above its long-term range of 4%–6% of revenue. That money goes into heads, media and factory automation to fit more capacity on each drive, not into building more drives. This supply discipline is central to the whole industry, and it's the main reason prices have kept rising.

The customer base is extremely concentrated. Cloud revenue was $3.87 billion this quarter, 88% of the total, up 54% year over year. Client was $320 million, up 119%. Consumer was $193 million, up 19% (10-Q, 2026-10-23; earnings call). By region, the Americas contributed $2.08 billion, Asia $1.71 billion, and Europe, the Middle East and Africa $592 million, roughly 47%, 39% and 14% respectively (10-Q, 2026-10-23). The filing doesn't disclose how much revenue comes from any single customer, but the phrase "top seven customers" makes the point: a handful of hyperscale cloud providers hold the company's fate in their hands.

Take the contract language with a grain of salt. "Visibility out to 2031" sounds like five years of locked-in orders, but management attached three caveats: LTAs cover a smaller share of volume the further out you go, some agreements are expiring and being renegotiated, and part of the shipments sit outside any contract and are priced on the spot market (earnings call, 2026-10-22). That's a very different thing from the remaining performance obligations (RPO) that cloud-software companies report. WDC doesn't disclose RPO. Its "visibility" is demand planning that customers share in advance, and neither the volume nor the price is a binding, non-cancellable commitment.

The weakest link is HAMR qualification timing. Rival Seagate's HAMR drives already make up about 40% of its nearline shipments (nearline drives are the high-capacity drives that fill data centers). WDC's HAMR won't finish qualification with two hyperscale customers until the first quarter of 2027, with volume production starting in the first half (earnings call, 2026-10-22). Until then, WDC relies on its ePMR platform, which tops out at 40TB per drive. If HAMR qualification slips, price increases become the only source of growth, and price is the least reliable variable in any cycle.

Few companies line up neatly against WDC. Revenue growth and margins are the two cleanest yardsticks:

Pure Storage's gross margin is higher because it sells all-flash storage systems bundled with software, a product with completely different value and cost economics from a bare hard drive. Its operating margin is only about four-tenths of WDC's. The only true comparison is Seagate. Together, the two companies make up nearly the entire nearline hard-drive market.

The two companies are taking very different paths. WDC ships more and has a slightly higher gross margin, but its shipment growth is only two-thirds of Seagate's, and the reason is the HAMR technology gap. Seagate sells more exabytes by packing more capacity into each drive, while WDC makes up the difference by raising prices. While prices keep rising, the gap barely shows. Once price increases slow, the company with the capacity lead will win market share first. Seagate hasn't reported this quarter yet. When it does, its shipment growth and what it says about pricing will be the most direct check on WDC.

The only storage peer to report this quarter so far is Micron. Micron reported after the close on 2026-09-30 and beat EPS estimates by 5.7%, driven by the same mix of surging demand and higher prices. Management called FY2026 a record and said it expects FY2027 to be stronger. Yet the stock rose only 3.0% the next day, less than half the ±6.2% move options traders had priced in (peer earnings-reaction data, 2026-10-22). A strong report and even stronger guidance earned only half the expected move. That tells you the market had already priced in its optimism about the storage pricing cycle, and what worries investors now is how long the peak can last. Treat that as a loose read-across for WDC, though. Over the past 8 quarters, the two stocks moved the same way after earnings only 2 times, with a correlation of -0.61, so Micron's reaction tells you little about how WDC will trade.

WDC's own options market priced in a ±7.5% move this time. In 6 of the past 8 quarters the actual move was bigger than the implied one, with a median implied move of ±6.9% against a median actual move of 8.4% (same source). This time the stock fell just 3.0% the next day, well inside the implied range. The market gave almost no reward for guidance that blew past expectations, and that in itself is a signal.

The price-to-earnings ratio still works for WDC, because its profits come from real operations and are backed by cash flow. The question is which period's earnings to use. The trailing four quarters understate earnings that are still rising, while annualizing peak earnings overstates how long the cycle will last. So we show both.

How cheap is 15x? The S&P 500 has long traded at around 20x forward earnings. Pricing a company growing revenue 55% with an operating margin near 50% at only 15x means the market isn't valuing WDC as a growth stock. It's valuing it as a cyclical, on the assumption that today's profits are close to a cycle peak and will fall back at some point. On 12x EV to operating profit, the stock isn't cheap if annualized profit holds for only about two years and then halves. If the pricing cycle lasts beyond 2028, as management has hinted, today's valuation looks clearly too low.

In other words, today's price reflects fear of a cycle peak, and the market is giving almost no credit for growth. Earnings are rising while the valuation multiple shrinks. That's what makes it possible for both earnings estimates and the multiple to rise together later. For the multiple to recover, though, the market first needs proof that price increases won't reverse within a quarter or two, and only the coming quarterly reports can provide that.

The hard-drive cycle doesn't move in lockstep with DRAM and NAND memory chips. It behaves more like a cycle run by a small group of producers who keep supply tight. Two companies control the nearline market, and for years neither has added drive-making capacity. They've grown exabyte supply only by fitting more capacity on each drive, while storage demand from AI training data, inference logs and video has grown faster. That gap between supply and demand shows up in prices. Three sets of readings tell us where we are:

Taken together, my view is that the hard-drive cycle is in the middle-to-late part of its upswing. It's mid-cycle because price increases are still accelerating, contracts are getting longer, and customers' spending forecasts are still being raised. It's on the late side because gross margin is already at 60%, a level the industry has never reached before, and consumers are starting to feel the cost squeeze. TrendForce noted on 2026-10-08 that rising costs will weigh on laptop demand in 2027. On top of that, Seagate's HAMR ramp and WDC's own HAMR production will both add a lot of exabyte supply in 2027.

Two signals could flag a turning point early:

This section covers week-to-week signals that bear on timing over the next few weeks. Readers who plan to hold for a year or more can skip it.

Money is leaving storage. As of 2026-10-23, the storage sub-sector has spent 10 straight days in the weakest money-flow tier. On the relative-strength rotation chart it sits in the "lagging" quadrant, with a money-flow reading of -58 and 8 straight days of net outflows. It dropped out of the middle tier on 2026-10-12. Net flow over the past 5 days totals about -955 million shares, while the 20-day total is still +770 million shares (share volume estimated from price and trading volume, not dollars). So buyers were still coming in during the first half of the month, and the selling has bunched up in the past week. The sector is down 2.6% over 5 days (sector money-flow data, 2026-10-23). Against that backdrop, WDC's 3% drop after earnings looks more like collateral damage from selling across the sector than a verdict on the report itself.

One level up, the broader technology sector is also in the weakest tier, at the bottom alongside communication services. Energy and consumer staples are in the strongest tier, followed by materials, financials, real estate and utilities (2026-10-23). Where the money is going is clear. The rotation-strength reading from technology into materials is 2.54, into financials 2.09, and into real estate 1.32. The three sub-sectors with the biggest inflows are IT services (reading 70), electric utilities (66) and alcoholic beverages (65). The biggest outflows are from lithium batteries (-81), quantum computing (-80) and semiconductor etch and deposition equipment (-75) (rotation data, 2026-10-23). This is a textbook move into defensive stocks. The broad signal is risk-off, and the market-regime reading is -35, in bear-market territory. SPY closed at 759.97, about 1.32% below its 20-day moving average of 770.16. Liquidity is also getting tighter: net liquidity is $5.80 trillion, down 1.2% week over week, and the 10-year Treasury yield is 4.96% (2026-10-23).

Tech isn't moving as one block, though. IT services moved up a tier on 2026-10-20 and enterprise software on 2026-10-22, which shows money inside tech shifting from hardware and semiconductors toward software and services, and storage is on the selling side. Has the stock already fallen further than the outflows justify? WDC has dropped below both its 50-day and 200-day moving averages, falling faster than money has left, so a good deal of the selling has already happened. History isn't very encouraging, though. After entering the weakest tier, the median return over the next 20 trading days is 0.0%, and the stock rises 55% of the time (3074 cases, covering 2021-04-12 to 2026-09-24). That's basically a coin flip. Being in the weakest tier predicts neither a rebound nor a further fall. The useful signal is when it moves back up a tier.

Long term (quarters to years): worth following. The fundamentals are the foundation. Revenue is up 55%, operating margin is 49.1%, next-quarter guidance beats consensus by about 20%, free cash flow easily covers both buybacks and dividends, and there's almost no gap between GAAP and adjusted earnings. The valuation is what makes the case attractive: about 15x guided annualized earnings means the market is pricing a cycle peak and giving no premium for growth. Be clear about the costs, too. This is a highly volatile cyclical stock that swings more than the market (high beta). It had already fallen about 24% before earnings, and in 6 of the past 8 quarters the options market underestimated its post-earnings move, so drawdowns of 20% or more along the way are normal. HAMR is a generation behind Seagate, and price is the main driver of current growth, which is also the first thing to turn in a cycle. Our suggestion is to let the checkpoints guide you. The next report showing price per TB up at least 35% and gross margin above 60%, and HAMR finishing qualification at two hyperscale customers in Q1 2027, are three milestones. Each one passed justifies a bit more conviction. On the other hand, if price growth slows for two quarters in a row, treat that as a sign the cycle has peaked. Don't wait for profits to fall before acting.

Short term (days to weeks): don't chase it; wait for the selling to stop. The storage sub-sector has spent 10 straight days in the weakest tier with 8 straight days of net outflows, the broader tech sector is at the bottom, the overall market is risk-off with tightening liquidity, and the stock is below both of its major moving averages. In conditions like these, a strong report may not move the stock, as Micron already showed. Short-term traders need to wait for money flows to turn: specifically, net outflows from the sub-sector stop, the money-flow reading clearly recovers or moves up a tier, and the stock climbs back above its key moving averages. Until at least one of these happens, good fundamental news mostly hands long-term investors better entry prices. It isn't yet a reason for short-term money to jump in.

The second half turns these two calls into action: the two entry conditions to watch for, and three reference price levels built around them.