Western Digital’s nearline beat is being underpriced because cash conversion, not revenue, is the new signal

The first distinction matters because the headline revenue surprise was not the whole event. What was priced in was $2,929.4 million of revenue and $1.93 of EPS, effectively a quarter where demand improved but did not force a major reset in profitability. What actually surprised was cleaner: revenue landed at $3,017.0 million, a +3.0% surprise, while EPS landed at $2.13, a +10.4% surprise. That spread between the top-line beat and the EPS beat is the core of the thesis. Investors can debate whether nearline demand was expected, but they cannot ignore that the earnings sensitivity to each dollar of revenue was better than the Street estimate embedded. Management’s own reported framing also points to the same conclusion, though on its call basis rather than the Street-comparison basis: CFO Kris Sennesael said, “During the second quarter of fiscal 2026, revenue was $3 billion, up 25% year-over-year, driven by strong demand for our nearline drives.” The phrase matters less for the demand adjective than for the explicit product attribution: this is not a broad HDD recovery story, it is a nearline capacity story with unusually favorable mix.

That mix conclusion is defensible because the quarterly trajectory shows revenue recovering without giving back gross margin. Revenue moved from $2,605.0 million in Q4 FY2025 to $2,818.0 million in Q1 FY2026 and $3,017.0 million in Q2 FY2026, while gross margin moved from 41.0% to 43.5% to 45.7%. The sequential revenue growth rates, +13.6%, +8.2%, and +7.1%, did not require margin sacrifice; the opposite occurred. That is the market’s likely miss: HDD investors often fade revenue acceleration when they assume pricing or mix must normalize, but this print shows the margin line still rising. The historical comparison is also useful because the current quarter is not merely lapping a depressed base. Q2 FY2025 revenue was $4,285.0 million with 35.4% gross margin, meaning the company is now operating with a smaller revenue base than that spike but a materially higher gross-margin structure at 45.7%. The print is telling PMs not to anchor on peak revenue alone.

The capacity detail explains why revenue quality improved, not just why revenue beat. Sennesael disclosed, “We delivered 215 exabytes to our customers, up 22% year-over-year.” He also said the quarter included “over 3.5 million drives or 103 exabytes of our latest generation ePMR with capacity points up to 32 terabytes.” Those numbers put hard dimensions around the mix shift: 103 exabytes of latest-generation ePMR inside 215 exabytes shipped means the product cycle is already large enough to affect company economics, not merely an early qualification story. The cloud skew reinforces the same read. Management said cloud represented 89% of total revenue at $2.7 billion, while client represented 6% at $176 million and consumer represented 5% at $168 million. In practice, WDC has become a nearline-cloud earnings vehicle for this cycle, not a balanced storage hardware story. That should change how investors value the print: the right debate is not whether PC or consumer storage can help, but whether cloud nearline pricing and high-capacity adoption can preserve the current margin ramp.

The margin guide is the clearest forward evidence that management sees the same profit pool the quarter revealed. On the call, Sennesael stated that WDC delivered 46.1% gross margin, up 220 basis points quarter-over-quarter and up 770 basis points year-over-year, then guided to 47% to 48%, or 47.5% at the midpoint, up 740 basis points year-over-year. The reported history table shows Q2 FY2026 gross margin at 45.7%, so the call and table are on different presentation bases, but both point in the same direction: margins rose in Q2 and the next-quarter guide embeds another step up. The guidance also puts boundaries around the thesis. Revenue is expected to be $3.2 billion, plus/minus $100 million, operating expenses are expected in the range of $380 million to $390 million, and diluted EPS is expected to be $2.30, plus/minus $0.15, based on a non-GAAP diluted share count of approximately 385 million shares. If revenue were the only story, the EPS guide would matter less. Here it matters because OpEx growth is contained to a stated $380 million to $390 million band while gross margin is guided higher.

The balance sheet and capital return details make the variant perception more actionable because the company is converting the cycle into distributable cash, not just accounting EPS. Operating cash flow was $745 million, capital expenditures were $92 million, and free cash flow was $653 million, with a free cash flow margin of 21.6%. Cash and cash equivalents were $2 billion, total liquidity was $3.2 billion including the undrawn revolver capacity, debt outstanding was $4.7 billion, net debt was $2.7 billion, and net leverage EBITDA was well below 1 turn. The market may discount HDD earnings as cyclical, but the company is reducing that objection by returning capital while keeping leverage contained. During the quarter, WDC paid $48 million of dividends and repurchased $615 million of stock, buying 3.8 million shares of common stock. Since the program began in Q4 FY2025, it has returned $1.4 billion through repurchases and dividends. Sennesael’s wording on buybacks was unusually direct: “We already have repurchased $1.3 billion or we have used $1.3 billion of that program, repurchasing on or about 13 million shares, and there is no hesitation.” That language matters because it removes some ambiguity over whether capital return pauses if management sees a cyclical peak.

That confidence does not eliminate the capacity-cycle risk, but the CapEx comments make the bearish version harder to prove from this data pack. Tiang Yew Tan said UltraSMR gives a 20% capacity uplift over CMR and a 10% capacity uplift over standard industry SMR, which means WDC is trying to satisfy exabyte demand through areal-density gains rather than a simple equipment-spend surge. He also said that even with the HAMR ramp expected at the start of calendar year ’27, CapEx as a percentage of revenue on a run-rate basis will still be within the 4% to 6% range. That is a meaningful constraint for the bear case. If WDC were signaling a large investment cycle to chase nearline demand, investors would be right to haircut free cash flow. Instead, the company paired $92 million of capital expenditures in Q2 with a stated 4% to 6% run-rate CapEx framework even through HAMR. The thesis can break if that framework moves higher, but the current evidence says the earnings cycle is not yet being consumed by capacity spending.

The tone of the call supports the financial read, though it is not a cleanly euphoric transcript. In the tone history, Q2 FY2026 sentiment was 0.35, down from 0.41 in Q1 FY2026, while guidance_tone stayed flat at 0.32. Prepared_sentiment fell to 0.40 from 0.64, qa_sentiment eased to 0.34 from 0.37, uncertainty rose to 64.8 from 59.1, and qa_evasiveness increased to 15.7 from 3.2. Those numbers conflict with the fundamental acceleration, and the conflict matters: management delivered a beat and a higher-margin guide, but the transcript was less celebratory and more uncertain than the previous quarter. I would not read that as a bearish reversal, because tone_confidence increased to 0.51 from 0.39, suggesting the language was measured rather than confused. Still, the call delivery argues against paying for an unchecked cycle. The right stance is constructive, but with confirmation required in the next guide.

The read-through to the rest of the storage chain is narrower than usual because the supply-chain data pack names no customers of WDC and no suppliers to WDC. That absence is itself useful discipline: this print should not be stretched into a named hyperscaler, component vendor, or equipment read-through without evidence here. The only customer set quantified by the company is category-level cloud at 89% of total revenue and $2.7 billion, with client at 6% and $176 million and consumer at 5% and $168 million. The implication is therefore for cloud storage procurement budgets broadly rather than for a named customer account. On the supplier side, the $92 million of capital expenditures and 4% to 6% CapEx-as-revenue framework limit the immediate upside read-through to unnamed equipment and capacity suppliers. For portfolio construction, that means WDC’s print is a cleaner signal for HDD nearline pricing and mix than for a multi-company semiconductor capital equipment basket.

The peer comparison also argues that WDC’s opportunity is being measured against the wrong cohort if investors lump it with memory names showing extreme cyclical rebounds. In the latest reported quarter peer set, WDC’s $3,337.0 million of revenue, 50.2% gross margin, and +45.5% revenue YoY sit much closer to STX at $3,112.0 million of revenue, 46.5% gross margin, and +44.1% revenue YoY than to MU at $41,456.0 million of revenue, 84.6% gross margin, and +345.7% revenue YoY or SNDK at $5,950.0 million of revenue, 78.4% gross margin, and +251.0% revenue YoY. The actionable point is not that WDC deserves the same multiple as higher-margin memory peers. It is that WDC’s HDD profit pool is improving in a more controlled way, with revenue YoY of +45.5% in the peer table and gross margin of 50.2% in the subsequent quarter shown in the history. Against STX, the comparison is direct and favorable on the numbers provided: WDC is ahead on revenue, gross margin, and revenue YoY in the latest peer table.

The other historical nuance is that Q3 FY2026 in the quarterly history already shows what the Q2 guide was trying to set up: revenue of $3,337.0 million, gross margin of 50.2%, revenue QoQ of +10.6%, revenue YoY of +45.5%, and diluted EPS of $8.52. Because the earnings event under discussion is Q2 FY2026, those Q3 FY2026 figures should be treated as trajectory evidence in the data pack rather than as part of the Q2 beat. They nonetheless sharpen the thesis: the company’s Q2 guide to $3.2 billion, plus/minus $100 million, and 47% to 48% gross margin was not pointing to a fade. It was pointing to another leg higher. The market’s mistake would be to focus only on Q2 revenue being below Q2 FY2025’s $4,285.0 million, while missing that gross margin has moved from 35.4% in Q2 FY2025 to 45.7% in Q2 FY2026 and then to 50.2% in Q3 FY2026 in the history.

What to watch next is therefore specific. The thesis is confirmed if next-quarter commentary sustains revenue at or above the guided $3.2 billion midpoint, stays within the $3.2 billion plus/minus $100 million range, and holds gross margin inside or above the guided 47% to 48% range. It is reinforced if operating expenses remain within $380 million to $390 million, EPS stays within or above $2.30 plus/minus $0.15 on approximately 385 million shares, and capital expenditures remain consistent with the 4% to 6% run-rate framework. It starts to break if cloud falls materially below 89% of total revenue or $2.7 billion, if latest-generation ePMR shipment color retreats from the Q2 level of over 3.5 million drives and 103 exabytes, or if free cash flow fails to track the Q2 benchmark of $653 million and 21.6% free cash flow margin. The next dated capital-return marker is the quarterly cash dividend of $0.125 per share payable on March 18, 2026 to holders of record as of March 5, 2026; the more important investor test will be whether buybacks continue after $615 million in Q2 repurchases and $1.4 billion returned since Q4 FY2025.