Toshiba Breaks a Five-Year Silence on HDD Capacity
TREE NEWS reports: Toshiba announced a ¥60 billion (roughly $380 million) investment to expand its hard disk drive plant in the Philippines, with plans to double capacity by fiscal 2027 to capture incremental demand from AI data centers. The news triggered an immediate sell-off in the two other members of the HDD oligopoly: Seagate (STX) and Western Digital (WDC) each plunged more than 10% on Friday — the most conspicuous losers on a day when the Nasdaq Composite closed at a record high.
The context makes the reaction intelligible. Seagate is up 208% year-to-date and Western Digital 141%, versus a 22% gain for the Nasdaq 100. The entire rally rested on a single premise: nobody is building new factories. Now someone is.
Why ‘Supply Discipline’ Was the Whole Thesis
The HDD upcycle — and, by extension, the commodity semiconductor chain spanning DRAM and NAND — has been underwritten by supply discipline. All three drive makers refrained from adding drive capacity, letting all volume growth come from areal density improvements. In August, after hosting Seagate management at a non-deal roadshow, Bernstein wrote explicitly that Seagate, Western Digital and Toshiba were all holding capacity flat, keeping industry supply structurally disciplined. Seagate management said its factories were running full and that it would not expand, instead lifting nearline exabyte shipments at roughly a 25% CAGR through HAMR technology. New fabs, management noted, take at least two years — and there were no such plans. Bernstein promptly named Seagate a top pick on precisely that logic.
The tightness is real: Seagate said in January that its entire 2026 nearline capacity was sold out. On Friday, Toshiba — which holds a little over 10% of global capacity — said it is adding supply.
Timing Is Everything
The timing is telling. Rosenblatt noted that hyperscalers are negotiating long-term agreements stretching into 2029–2031, and Toshiba’s expansion “introduces a credible medium-term supply risk, giving customers additional negotiating leverage.” Evercore added that Seagate has locked most of its nearline capacity through 2028, while Western Digital is still negotiating LTAs extending to 2031. In other words, 2027–2028 pricing is largely locked; what the market is repricing is the far end of the curve — and for stocks valued on forward earnings power, that is precisely the part that matters most.
Sell-Side Rushes to Defend
The uniformly bullish sell-side quickly pushed back. Citi argued the bottleneck is not Toshiba’s plant but its suppliers: unlike Seagate and Western Digital, Toshiba does not make its own media and heads, so doubling EB capacity would require its external component suppliers to expand substantially — which Citi expects to limit the actual incremental supply. Morgan Stanley said the HDD supply-demand gap through 2028 remains larger than Toshiba’s planned addition, noting Toshiba lacks leading-edge capacity and HAMR technology. Rosenblatt characterized Toshiba’s real intent as recapturing lost share, with the 30% target more posture than plan. Bloomberg Intelligence framed the expansion as validation of strong AI storage demand rather than near-term supply pressure. Sell-side price targets now sit 50% to 90% above Friday’s close.
The HAMR Moat and the Real Question
Do the math: doubling from just over 10% share to roughly 20% adds about 10% of industry EB supply over two years — while Seagate alone is growing nearline EB at roughly 25% annually through HAMR. Toshiba’s plan is large in headline terms and close to a rounding error in EB terms, provided its head and media suppliers cooperate. More importantly, this is not a fair fight. Seagate is the only vendor shipping HAMR at scale, and its areal density lead has translated into faster EB growth. Bernstein forecasts Seagate’s gross margin will overtake Western Digital’s in fiscal 2027 and climb toward 65%. If Toshiba’s expansion lets the tide go out, the vendor with the weakest areal density roadmap is most exposed — and Western Digital has indeed fallen harder than Seagate since June highs.
Many analysts believe Toshiba’s ¥60 billion will not dent HDD pricing before 2028: capacity is sold, LTAs are signed, and Toshiba still must buy core components. On Friday, HDD equipment maker Veeco rose 11% against the tape — its ion beam deposition and etch tools make read-write heads, so it benefits whoever expands. But the market was never pricing just the next two years; it was pricing a decade of quasi-monopoly discipline among three oligopolists. In commodities, the cure for high prices is high prices — sooner or later, someone builds a factory.
Key Takeaways for Investors
- The thesis just got a crack. The HDD trade was a supply-discipline trade. Toshiba’s move introduces a credible medium-term supply risk and hands hyperscaler customers negotiating leverage in LTA talks extending to 2031.
- Near-term pricing is locked, long-end is not. With 2027–2028 capacity largely sold, the repricing is happening at the far end of the curve — exactly where forward-earnings multiples are most sensitive.
- Watch for a response. If Seagate and Western Digital announce their own expansions, the supply-discipline theme is over. If they hold, Toshiba is merely buying share at the cycle top.
- Quality within the group matters. HAMR leadership gives Seagate a margin and areal-density edge; Western Digital is more exposed if pricing power erodes. Equipment suppliers like Veeco benefit regardless of who expands.
