ACMR’s beat is not the story: the market should price the margin reset, not the revenue guide

The print says ACMR is still gaining revenue, but not yet earning the incremental dollar the way bulls need it to. What was priced in was a modest revenue beat against $241.3 million and a continuation of the company’s multi-year top-line compounding; what actually surprised was that $244.4 million of revenue, +1.3% versus the Street, still produced EPS of $0.25 versus $0.39, a -35.9% miss on the street-comparison basis. That separation matters because the revenue surprise was not enough to absorb a gross margin trough of 40.9%, down from 42.0% in Q3 FY2025 and below the 49.6% posted in Q4 FY2024. The market can rationalize a single quarter of margin pressure in wafer fab equipment when shipments are lumpy, but this was the second consecutive quarter below the long-term band management itself referenced, with Q3 FY2025 at 42.0% and Q4 FY2025 at 40.9%. The thesis, therefore, is not that demand is broken. It is that the market is too willing to capitalize the 2026 revenue range before confirming that the company can re-expand gross margin and fund a larger operating platform without repeating the EPS compression embedded in this print.

The revenue trajectory still gives bulls something real to underwrite, which is why the print is not simply negative. Q4 FY2025 revenue of $244.4 million was up +9.4% YoY, and the prior quarter’s $269.2 million was up +32.0% YoY, so the company is not facing the kind of top-line air pocket that would invalidate the long-term customer penetration story. But the sequential pattern is less forgiving: revenue fell -9.2% QoQ in Q4 FY2025 after rising +25.0% QoQ in Q3 FY2025 and +25.0% QoQ in Q2 FY2025. The right interpretation is that ACMR’s year-end order conversion was adequate relative to the Street’s $241.3 million, but not adequate relative to the cost base and gross margin assumptions embedded in the $0.39 EPS estimate. That is the variant perception: consensus appears to have modeled a cleaner revenue-to-EPS translation than the quarter delivered.

The gross margin line is where the quarter became investable or dangerous depending on one’s prior. Q4 FY2025 gross margin of 40.9% was below the company’s stated long-term target range of 42% to 48%, and David Wang’s wording matters because it concedes the breach without framing it as structural: “Q4 gross margin was slightly below our long-term target range of 42% to 48%.” The word “slightly” is doing work, but the history is more demanding than the adjective. Gross margin was 53.8% in Q1 FY2023, 52.5% in Q3 FY2023, 52.0% in Q1 FY2024, and 51.4% in Q3 FY2024 before moving to 47.9% in Q1 FY2025, 48.5% in Q2 FY2025, 42.0% in Q3 FY2025, and 40.9% in Q4 FY2025. Management can defend the quarter as just below the target range, but investors should require the next data point to prove the 42% to 48% band is still the right normalized framework. The Q1 FY2026 history entry is encouraging on that specific issue, with gross margin at 46.4%, but because the earnings event under review produced the Q4 FY2025 EPS miss, the burden has shifted from revenue growth to repeatable margin recovery.

That burden is heavier because the spending plan is not a maintenance plan. Mark McKechnie gave a compact 2026 cost architecture that deserves more attention than the revenue outlook: “For 2026, we plan for R&D in the 16% to 18% range, sales and marketing in the 7% to 8% range and G&A in the 6% range.” Against a 2026 revenue outlook of $1.08 billion to $1.175 billion, those ratios signal management is choosing product breadth, customer support, and corporate scale over near-term operating leverage. The full-year 2025 expense base already moved against earnings: operating expenses were $258.4 million, up 34%, while full-year revenue was $901.3 million, up 15.2% on the company’s own reported basis. Net income attributable to ACM Research was $110.2 million versus $152.2 million, and full-year net income per diluted share was $1.61 versus $2.26. Those numbers are not a demand recession; they are a reinvestment cycle combined with margin compression. If the market prices the company only on the midpoint of the 2026 revenue range, it is missing that management has given explicit permission for elevated R&D, sales and marketing, and G&A consumption.

The capex plan makes the same point in capital rather than expense terms. Capital expenditures were $58 million for the full year 2025, and McKechnie said, “For the full year 2026, we expect to spend about $200 million in capital expenditures.” That step-up is not trivial in the context of Q4 FY2025 shipments of $228 million, down 13.5%, and full-year 2025 shipments of $854 million, down 12.2%. The conflict investors have to underwrite is clear: reported full-year 2025 revenue grew to $901.3 million, up 15.2%, but shipments declined $854 million, down 12.2%. That divergence can reflect timing and recognition mechanics rather than end-demand deterioration, but it still means the company is entering a materially larger capex year after a year when shipment dollars went backward. The $111 million gross proceeds from the February 2026 sale of approximately 4.8 million ACM Shanghai shares helps fund the buildout, as do the dividends net of tax of $19.2 million, $28.5 million and $29 million, respectively. But the investment case now relies on those funding sources translating into customer-qualified capacity, not merely a bigger balance sheet and bigger factories.

The product mix explains why management is willing to spend through the EPS disappointment, and also why investors should be selective in giving credit. Full-year 2025 revenue from single-wafer cleaning, Tahoe and semi-critical cleaning tools was $626 million, up 8%, and represented 69% of total revenue. That is still the core franchise and the biggest read-through for customers. Q4 revenue for single-wafer cleaning, Tahoe and semi-critical cleaning was $159.9 million, up 3%, which means the largest product group was growing but not at the pace of the company’s faster lines. Q4 revenue for ECP, Frontend Packaging, Furnace and other technologies was $64.1 million, up 23.9%, and Q4 revenue for Advanced Packaging, excluding ECP, services and spares was $20.5 million, up 23.8%. For the full year, advanced packaging, excluding ECP but including service and spare, was up 45% to $76 million and represented 8% of revenue. The strategic message is that growth is increasingly coming from categories outside the largest cleaning bucket, but the margin message is unresolved because Q4 gross margin was 40.9% while those growth categories were expanding faster than the core. Until the company shows that the higher-growth mix can coexist with gross margin inside 42% to 48%, the faster categories should be valued as revenue optionality, not yet as earnings accretion.

The customer concentration data reinforces the same underwriting discipline. In 2025, ACMR had 4 customers above 10% of sales, with the top customer at 16.9%, the next at 13.5%, then 11.6% and 10.2%, for an aggregate total of 52.2% of total sales. For SMIC, which uses ACMR’s single-wafer cleaning products including SAPS and TEBO, the implication from this print is not that Chinese domestic tool demand is weakening. The relevant magnitude is that ACMR’s cleaning-related full-year revenue was $626 million, up 8%, and Q4 cleaning-related revenue was $159.9 million, up 3%, while the customer base remained concentrated enough that 4 customers accounted for 52.2% of sales. That combination says leading customers are still absorbing tools, but the pace in the core cleaning bucket is slower than the +23.9% Q4 growth in ECP, Frontend Packaging, Furnace and other technologies and the +23.8% Q4 growth in Advanced Packaging, excluding ECP, services and spares. The second-order read-through for SMIC is therefore measured: continued spend in single-wafer cleaning is supported by ACMR’s $626 million full-year category size and +8% growth, but incremental ACMR growth is tilting toward adjacent process areas rather than a pure acceleration in cleaning.

The competitive comparison cuts both ways, which is why the multiple should not be rewarded simply for being in wafer fab equipment. ACMR’s Q4 FY2025 revenue YoY growth of +9.4% sits below TOELY’s +10.6% and below 6361.T’s +15.8%, while it is above 7735.T’s +9.1%, 7751.T’s +3.3%, 6302.T’s 0.0%, 7731.T’s -4.6%, and 6525.T’s -3.5%. On gross margin, ACMR’s 40.9% was below 7751.T at 46.2% and TOELY at 46.8%, above 6361.T at 31.6%, 6302.T at 25.0%, 7731.T at 40.5%, 7735.T at 40.8%, 6728.T at 32.0%, and 6525.T at 39.4%. The peer lens says ACMR is not a laggard on growth or margin, but it also does not deserve an unqualified scarcity premium when two larger Japanese-listed peers in the table show gross margins of 46.2% and 46.8% and one peer shows revenue YoY of +15.8%. The cleaner relative argument is that ACMR offers a differentiated China and product-expansion growth path, but the Q4 EPS miss means that path is still carrying higher execution cost than investors modeled.

The call delivery was better than the P&L, and that is useful but not sufficient. The tone history shows Q4 FY2025 sentiment at 0.41, up from 0.27 in Q3 FY2025, guidance_tone at 0.26 versus 0.11, tone_confidence at 0.65 versus 0.52, prepared_sentiment flat at 0.53, qa_sentiment at 0.29 versus 0.00, ai_optimism at 0.52 versus 0.36, uncertainty at 94.0 versus 95.9, and qa_evasiveness at 18.0 versus 40.5. That is the profile of a management team that answered with more specificity and less avoidance even as the margin and EPS optics worsened. The later Q1 FY2026 tone series extends that improvement, with sentiment at 0.54, guidance_tone at 0.70, qa_sentiment at 0.58, uncertainty at 44.0, and qa_evasiveness at -55.3; the call-over-call delta from Q4 FY2025 to Q1 FY2026 was sentiment +0.14, guidance_tone +0.44, qa_sentiment +0.28, uncertainty -50.0, and qa_evasiveness -73.2. The conflict is important: Q4 FY2025 fundamentals asked for skepticism, while subsequent call delivery became more direct. I would treat that as a reason not to short the guide mechanically, not as a reason to ignore the margin reset.

Management’s own ambition is explicit enough that the market will keep looking through quarterly volatility if gross margin cooperates. David Wang said, “We have the customer, the product, the capacity and the capital to execute our global business plan, and we remain committed to our long-term target of $4 billion in revenue.” That quote earns attention because it ties the $200 million 2026 capex plan, the product expansion, the customer concentration, and the $111 million gross proceeds into one strategic claim. But the numbers around it impose a high bar. Full-year 2025 revenue was $901 million in Wang’s framing and $901.3 million in McKechnie’s framing, up 15.2% on the latter’s basis, while full-year net income attributable to ACM Research fell to $110.2 million versus $152.2 million and full-year net income per diluted share fell to $1.61 versus $2.26. A long-term $4 billion target is not the debate for the next quarter. The debate is whether the company can show that the route to that target does not structurally dilute gross margin and earnings power.

What to watch next is therefore narrow and numerical. The first confirmation point is gross margin: Q1 FY2026 history shows 46.4%, and investors should require the next quarter to remain inside the 42% to 48% target range rather than revisiting Q4 FY2025’s 40.9%. The second is the revenue bridge to the 2026 outlook of $1.08 billion to $1.175 billion; a quarterly revenue print materially below the Q1 FY2026 history level of $231.3 million would make the full-year ramp more dependent on second-half conversion. The third is spending discipline against management’s 2026 framework of R&D at 16% to 18% of revenue, sales and marketing at 7% to 8%, and G&A at 6%, because Q4 FY2025 operating expenses of $70.6 million, up 21%, already showed how quickly EPS can undershoot when gross margin is low. The fourth is shipment conversion after Q4 shipments of $228 million, down 13.5%, and full-year shipments of $854 million, down 12.2%; shipment growth would support the capex thesis, while another decline would make the $200 million 2026 capex plan look early. The fifth is customer breadth: the 4 customers at 16.9%, 13.5%, 11.6%, and 10.2% accounted for 52.2% of 2025 sales, so any reduction in concentration while holding cleaning revenue growth near the 2025 level of +8% would strengthen the durability case. The stock deserves credit for demand if those numbers move together, but after this print it should not get paid twice for a revenue guide and a margin recovery until the recovery is visible in the next quarter’s gross margin and EPS.