ULVAC’s revenue beat is real, but the margin miss says this is still a mix-cycle, not an earnings-cycle

This print means ULVAC has demand visibility, not yet earnings leverage. What was priced in was a subdued quarter: the street had revenue at ¥63,229.5 million and EPS at ¥104.89, effectively expecting sales around the prior quarter’s ¥52,729.0 million base to recover but not fully normalize. What actually surprised was the shape of the beat. Revenue came in at ¥71,164.0 million, +12.5% versus estimate, but diluted EPS came in at ¥92.19 against ¥104.89, a -12.1% surprise. That is the variant perception: the stock can be right to react positively to bookings and sales recovery, but wrong if it capitalizes this as a higher-quality margin cycle. The company delivered +35.0% revenue QoQ in Q2 FY2025 after -16.9% in Q1 FY2025, while gross margin slipped from 29.7% to 29.1%. In semiconductor equipment, the first derivative of revenue often gets paid; here, the first derivative of profit conversion is the constraint.

The separation between demand and conversion is unusually clean because the revenue beat sits against a weaker historical profit profile. Q2 FY2025 revenue of ¥71,164.0 million was close to the Q2 FY2024 level of ¥73,853.0 million, but revenue YoY was still -3.6%, and EPS was ¥92.24 versus ¥135.60 in Q2 FY2024. That tells the portfolio question more clearly than the headline beat: revenue recovered sequentially, but the year-on-year earnings base did not. The margin line makes the same point. Gross margin at 29.1% was below Q1 FY2025 at 29.7%, below Q4 FY2024 at 30.5%, below Q2 FY2024 at 32.2%, and below Q3 FY2024 at 33.3%. A bull can point to ¥71,164.0 million revenue and +35.0% QoQ; a PM should ask why that did not carry gross margin back above 30.5%. Until the answer is mix, utilization, or pricing with a visible path to reversal, the print is a demand confirmation with an earnings-quality discount.

The financial trajectory also argues against extrapolating a single quarter of shipment recovery into a linear upcycle. ULVAC’s revenue history is lumpy: ¥75,908.0 million in Q4 FY2023, ¥61,019.0 million in Q1 FY2024, ¥73,853.0 million in Q2 FY2024, ¥52,854.0 million in Q3 FY2024, ¥63,458.0 million in Q4 FY2024, ¥52,729.0 million in Q1 FY2025, and ¥71,164.0 million in Q2 FY2025. The Q2 FY2025 rebound looks large because it follows a weak Q1 FY2025, not because it sets a new high. The Q3 FY2025 line in the data pack then shows ¥67,738.0 million revenue, -4.8% QoQ, with gross margin improving to 32.0% and revenue YoY at +28.2%. That later gross-margin repair is important for the thesis, but it does not erase the Q2 message: demand is present, yet quarterly shipment mix is still driving the earnings outcome. A stock that prices this as smooth capex beta risks paying for the wrong variable.

The company’s own filings reinforce the same distinction because orders are healthier than sales, but profits are still catching down year on year. The call excerpts show six-month orders rising, with the company stating, “Orders received 116,378 137,105 17.8%.” That language matters because it gives the cleanest company-reported demand datapoint in the pack: orders were up 17.8% to 137,105 million yen, while the same excerpt says “net sales decreased by 10,979 million yen (8.1%).” On the profit side, operating profit decreased by 6,888 million yen (44.9%) to 8,456 million yen, ordinary profit decreased by 6,900 million yen (43.0%) to 9,154 million yen, and profit attributable to owners of parent decreased by 4,214 million yen (40.5%) to 6,200 million yen. These are not street-comparison figures, so they should not be mixed with the ¥71,164.0 million and ¥92.19 surprise basis; they are still useful because they show that the company’s own six-month accounts have stronger orders than earnings. The stock debate should therefore shift from “is capex coming back?” to “which orders convert at 29.1%, 32.0%, or better?”

That conversion question is also where the segment data becomes more useful than the headline beat. In the company excerpts, one segment shows orders received of 91,484 and 108,355, or 18.4%, with backlogs of 114,634 million yen, net sales of 95,870 million yen, and operating profit of 6,833 million. Another segment shows orders received of 24,894 and 28,750, or 15.5%, with net sales of 25,657 and 28,023, or 9.2%, and backlogs of 19,086 million yen, net sales of 28,023 million yen, and operating profit of 1,550 million. The second-order implication is that the equipment recovery is not evenly monetized: the larger segment carries 114,634 million yen of backlog and 6,833 million yen of operating profit, while the smaller segment carries 19,086 million yen of backlog and 1,550 million yen of operating profit. If investors are buying the +17.8% order number, they also need to underwrite which backlog bucket turns into revenue and at what gross margin. That is the fulcrum between a sales beat and a profit beat.

The supply-chain read-through is consequently positive for customer capex activity but not a clean positive for ULVAC’s earnings mix. The named customers are TSMC, for equipment and high-purity sputtering targets including Co, Ti, and W, and Samsung for sputtering targets. Orders up 17.8% to 137,105 million yen and the larger segment backlog of 114,634 million yen indicate procurement activity that matters to customers using vacuum equipment and sputtering-related materials. For TSMC, the read-through is that equipment and high-purity sputtering target demand is being placed in ULVAC’s book even while ULVAC’s net sales decreased by 10,979 million yen (8.1%) on the company’s six-month basis. For Samsung, the sputtering target read-through is similarly that the order environment is not frozen, given orders received of 24,894 and 28,750, or 15.5%, in the smaller segment. But because gross margin was 29.1% in Q2 FY2025 and only later shows 32.0% in Q3 FY2025, the customer signal is stronger than the supplier earnings signal. ULVAC’s customers may be spending; ULVAC has not yet shown that every yen of that spending lands at the margin investors want.

The peer frame makes the same point in relative terms: ULVAC’s latest reported quarter in the peer table has the highest revenue YoY at +28.2% among the listed wafer-fab-equipment set, but not the best margin. The reporter’s ¥67,738.0 million revenue and 32.0% gross margin compare with TOELY at ¥724,894.9 million revenue, 46.8% gross margin, and +10.6% revenue YoY; 7751.T at ¥1,093,653.0 million revenue, 46.2% gross margin, and +3.3% revenue YoY; and 6361.T at ¥246,311.0 million revenue, 31.6% gross margin, and +15.8% revenue YoY. That is the comparative tension: ULVAC has the better YoY growth line in the peer table, but its 32.0% gross margin is far below TOELY’s 46.8% and 7751.T’s 46.2%, while only slightly above 6361.T’s 31.6%. A valuation argument built on growth alone ignores that wafer-fab-equipment peers with lower YoY growth are reporting much higher gross margins. The print narrows the gap on demand, not on profitability.

The call delivery adds another reason not to overread the revenue beat, because management tone improved on guidance versus Q1 FY2025 but deteriorated sharply versus the immediately prior call sequence in the tone table. The tone history shows Q2 FY2025 sentiment at -0.10 versus Q1 FY2025 at -0.22, and guidance_tone at 0.18 versus -0.29, so the event itself was less negative than the previous quarter on those measures. But the later Q2 FY2026 call-over-call delta versus Q1 FY2026 shows sentiment -0.61 and guidance_tone -0.14, while uncertainty was still 31.7 after 32.9. That conflict matters: backward-looking financial delivery shows revenue surprise of +12.5%, while the language series does not show a clean decline in uncertainty. The most constructive interpretation is that management has better demand signals than it had in Q1 FY2025; the less constructive interpretation is that those signals remain hard to translate into guided earnings. The tone data supports neither panic nor a blank-check re-rating.

The balance-sheet and company-guidance excerpts are consistent with that measured stance rather than a more aggressive call. The company excerpt gives total assets of 382,624 and equity of 232,771 with 59.0 at December 31, 2025, compared with 375,062, 231,080, and 59.6 at June 30, 2025. It also references ¥225,758 million as of December 31, 2025 and ¥223,465 million as of June 30, 2025. These figures do not create a liquidity alarm; they do show that the core debate is operating conversion, not balance-sheet survivability. The guidance excerpt for June 30, 2026 gives 250,000, (0.5), 28,500, 7.5, 28,500, (0.4), 20,000, 19.9, and 406.33. The point is not to rebase every model line from a fragmented excerpt, but to note that the company is not describing an unconstrained profit surge in the same document that shows operating profit down 44.9% to 8,456 million yen on the six-month basis. The print should be bought only if the investor believes the margin trough is passing, not merely because the revenue estimate was low.

The short-term stock setup is therefore asymmetric in a more nuanced way than the headline suggests. If the market came in focused on whether ULVAC could beat revenue after Q1 FY2025 revenue of ¥52,729.0 million and EPS of ¥33.75, the answer is yes: Q2 FY2025 delivered ¥71,164.0 million revenue and EPS of ¥92.24 in the quarterly history, with the street-comparison EPS at ¥92.19. But if the market came in expecting a cleaner operating inflection after the prior year’s Q2 FY2024 revenue of ¥73,853.0 million, gross margin of 32.2%, and EPS of ¥135.60, the answer is no: Q2 FY2025 revenue was still -3.6% YoY, gross margin was 29.1%, and EPS was well below ¥135.60. That is why the stock reaction should be judged against what investors were actually paying for. A relief rally on the +12.5% revenue surprise is defensible; a sustained move that assumes peer-like margin convergence toward 46.8% or 46.2% is not supported by this data pack.

What to watch next is concrete. The thesis is confirmed if the next reported quarter sustains revenue near the Q3 FY2025 level of ¥67,738.0 million while holding gross margin at 32.0% or better, because that would show the Q2 FY2025 margin of 29.1% was mix timing rather than structural leakage. It is also confirmed if orders remain close to the six-month 137,105 million yen level, if the larger segment backlog around 114,634 million yen begins converting without a return to 29.1% gross margin, and if EPS moves away from the Q3 FY2025 ¥59.20 trough rather than repeating the Q2 FY2025 street miss pattern of -12.1%. The thesis breaks if revenue falls back toward Q1 FY2025’s ¥52,729.0 million, if gross margin returns below 30.5% after the Q3 FY2025 32.0% print, or if the next tone update shows uncertainty moving back toward 32.9 with guidance_tone below 0.01. The date to anchor is the next quarterly release after this 2026-02-10 event: investors need evidence that the +17.8% order growth to 137,105 million yen is turning into revenue at 32.0% gross margin, not another quarter where the top line beats and EPS disappoints.