Western Digital's Boom Is Real — and Already Largely in the Price

The stock is up about 160% year to date, and on the surface the reason looks close to unassailable. After spinning off its flash-memory business, Western Digital is now a pure-play maker of the giant hard drives that AI data centers buy by the petabyte. It says capacity for calendar 2026 is already sold out, backed by firm orders, with some multi-year contracts stretched into 2027 and 2028. Cloud customers are roughly 89% of revenue. The last quarter, reported in early August, showed revenue up 44% from a year ago at $3.75 billion, gross margin near 54%, and free cash flow of $1.28 billion.

That is a genuine operating inflection, and any factor screen should reward it. But a stock can be a great business and a poor entry at the same time, which is the question here. The same facts that made the story real are, to a large degree, already sitting in the price. My job is to check whether the factor stack that drove the run is still intact or whether valuation has caught up and the rest is coasting.

The company really did change. Western Digital no longer has to split its attention between hard drives and flash. Management is framing the future around high-capacity drives for AI data centers, where the price gap between storage technologies has swung hard in its favor: hard disk drive prices are up roughly 46% since September 2025, and solid-state storage now costs more than 16 times as much per byte for equivalent capacity. That is what pushes bulk data into spinning disks rather than flash, and it is why hyperscalers are locking in supply years ahead instead of ordering quarter to quarter.

The numbers back the shift. Revenue grew about 36% over the trailing year, the gross margin sits near 45% on a trailing basis and hit 54% in the most recent quarter, and operating margin is around 30%. Return on invested capital is roughly 39%, and free cash flow is up 174% from a year ago. On safety, this is the strongest Western Digital has looked in years: it carries negative net debt — about $0.5 billion more cash than total debt — and a debt-to-equity ratio of roughly 0.12. Growth, profitability, and balance-sheet safety all grade at the top of the stack. On those three factors alone, there is nothing stale about this story.

But valuation has already absorbed the good news. The market now prices Western Digital at roughly a $161 billion market capitalization against $447 a share. Put that against what the company itself says it can deliver. Management has framed a target of $20-plus earnings per share and over 50% gross margins as a "floor" for the next three to five years. At $447, even $20 of EPS from the top of management's own range implies a multiple around 22 times earnings — and that is the floor case they are counting on years out, not next year. On 2028 profit projections that outside analyses put near $2.2 billion, the current market cap works out to more than 70 times that earnings level. Against the $3.5 billion of trailing free cash flow, the market cap is about 46 times cash flow, a yield of barely 2%.

Cheapness, as always, is a sector question, and here the comparison is a cold shower rather than a rescue. Western Digital is the least expensive member of the storage group on trailing earnings — about 17 times versus Seagate's 59 and Micron's 22. But look at what the market is paying for the year ahead and the whole pack is priced for prosperity: Western Digital at roughly 88 times forward earnings, Seagate near 129, Micron near 155. Being the least expensive stock in a group that has collectively decided the boom is permanent is not a value signal; it means even the "cheap" name has had its decade of upside pulled into today's price. And that re-rating concentrates risk in a handful of buyers: when a few hyperscalers fund the sold-out backlog, a shift in how even one of them stores data would hit the projections those multiples depend on.

The timing layer no longer confirms the story. Momentum was the tool that flagged the explosion, and it has cooled. The stock is down more than 40% from its 52-week high, recently slipped below its 50-day moving average, and trades below where the moving averages were positioning it. Nothing about the contracted backlog has deteriorated — this is a re-rating, not a fundamental break — but for a system that treats momentum as a timing tool rather than a thesis, the signal has gone from confirmation to caution. That is the difference between the fundamentals saying "durable" and the price saying "expensive and correcting."

What a holder and a watcher should each take from this is not the same thing. For an existing position, the operational improvement justifies holding through the pullback; the balance sheet and booked demand are the reason a major winner is not cut mechanically just because the multiple got warm. For someone without a position, the burden has shifted the other way: you are being asked to pay ~22 times the company's own stated earnings floor and roughly 46 times trailing cash flow for a name whose near-term momentum has turned against it. The reward for being right at this price is thinner than it was a thousand dollars of run-up ago.

The realistic trigger that changes the risk/reward is estimates doing the catching up — consensus earnings continuing to ratchet higher as the contracted volumes convert into reported profit, which would make today's forward multiple look less extreme. Analyst estimates have been rising and the company keeps beating them, so that path is alive. AInvest's aggregate signal still labels the stock Buy, and that is a fair grade for the operating picture. But an aggregate label does not rewrite the arithmetic between the current price and the current facts. The business is stronger than it has been in a generation; the price has already charged ahead of most of that improvement, and the factor stack's timing layer is no longer cheering it on. That is not a sell thesis against a fully booked 2028. It is a reminder that great businesses and good entry prices are not the same thing, and today they are not both in this one.