ASML: One Excuse After Another
ASML’s (NASDAQ: ASML) second-quarter 2025 results were the latest installment in how the company frames its challenges—and chooses its scapegoats.
When former ASML President and Chief Excecutive Officer Peter Wennink was in charge, he leaned on supply chain disruptions to excuse underperformance. ASML’s current Chief Executive Officer and President Christopher Fouquet now invokes the looming shadow of tariffs and macroeconomic uncertainty to explain what is, at its core, a softening of demand at the very heart of semiconductor scaling.
On the surface, ASML still wears the crown of technological supremacy. But beneath the polished veneer of its press releases, cracks are spreading. Investors see a company that has spent over €10 billion developing High‑NA EUV (Extreme Ultraviolet), while its flagship EUV platform is already slowing.
In Q2 2025, ASML reported net sales of €7.692 billion, down slightly from €7.742 billion in Q1. Gross margin edged down to 53.7%, net income softened to €2.29 billion, and new lithography system shipments fell to 67 units, down from 73 units in the prior quarter. Net bookings climbed to €5.5 billion, up from Q1’s weak €3.9 billion, but not at a level that signals the AI-fueled resurgence ASML has hinted at. Even more telling were the EUV figures. In Q1, ASML shipped 14 EUV systems. In Q2, that dropped to 11 EUV systems. At the cutting edge of chipmaking, where Fouquet claims the most momentum, demand is already faltering.
Regional exposure adds another layer of ambiguity. China still accounted for 27% of system sales in Q2, flat from Q1. But the numbers no longer tell the full story. ASML halted disclosure of DUV system sales into China in January this year, following increasing U.S. pressure. Investors are now left guessing whether that 27% reflects sustainable demand or merely the trailing echo of pre-ban pull-ins.
China Demand: A Structural Ceiling
The China data tells a larger truth about ASML’s future revenue ceiling. For much of 2023 and early 2024, China backstopped global demand. Chinese fabs, anticipating looming Dutch export restrictions, rushed to acquire as many deep ultraviolet systems as possible. By Q2 2024, China accounted for nearly half of ASML’s system sales. But by Q4, it had collapsed to 27% and stayed there ever since. The same plateau is visible across the equipment landscape.
Table 1 captures this structural demand shift. It shows how every major WFE supplier—not just ASML—has watched its China share peak and then flatten. Applied Materials, Lam Research, KLA, and Tokyo Electron all mirror ASML’s trajectory. Without the artificial boost from Chinese pull-ins, the weakness in the global lithography market is now fully exposed.
What was once a growth engine has now become a cap on addressable market. And that ceiling will not lift until export controls are relaxed, which is unlikely given the geopolitical trajectory.
Lithography Units: A Telling Decline
China aside total unit shipments paint a clear picture of stagnation. In Q1, ASML shipped 73 new systems. In Q2, it shipped 67. Used systems rose from 4 to 9 units. The nuance here matters. Customers are leaning more on legacy capacity, stretching existing tools rather than investing in new high-priced systems.
What’s more, demand for high-end nodes is weakening due to inventory digestion, elongated product development cycles, and conservative customer spending amid global macro uncertainty. Hyperscalers have pulled back on aggressive expansion plans, especially after front-loading AI capex in 2023-2024. Meanwhile, smartphone and PC refresh cycles remain sluggish, leaving chipmakers with less incentive to accelerate node transitions.
The numbers reveal a softening core market, especially the uptick in used sales. It is cheaper to refurbish and redeploy than commit to fresh purchases, especially when the demand outlook for cutting-edge nodes is cloudy.
The TSMC Paradox
TSMC continues to expose the fragility of ASML’s narrative. The foundry is moving ahead with its N2 node entirely on Low‑NA EUV and advanced multi-patterning. It has not purchased any High‑NA systems. This proves that even the world’s most advanced fab can achieve 2nm performance without the €350 million-per-tool platform Fouquet insists will unlock the future. If TSMC can avoid High‑NA, then High‑NA is not yet indispensable.
The Intel Paradox
Intel (NASDAQ: INTC) has gone in the opposite direction—at least in public relations. It was the first to install High‑NA EUV, buying two tools for its Oregon D1 fab and processing 30,000 wafers through them in 2024. Intel positioned this as proof it was reclaiming process leadership.
But then came the contradiction. Intel outsourced its 2nm Nova Lake CPUs to TSMC. The outsourcing decision says more than any marketing campaign. Even with privileged early access to High‑NA, Intel either lacked the confidence or the yield maturity to produce Nova Lake internally. For ASML investors, this is a troubling signal. If even Intel, the early adopter, won’t trust High‑NA for volume production, what does that say about broader demand? Until Intel ships a competitive, high-volume product made entirely on High‑NA, these tools remain expensive showcases, not revenue catalysts.
Peer Analysis: Who Holds the Risk?
The Q2 shockwave did not stop at ASML. On the day of its earnings release, ASML stock fell 10.6%. Applied Materials dropped 3.6%, Lam Research 3.0%, and KLA 2.6%. This synchronized drop reveals how interlocked the wafer-front-end ecosystem is. Lithography drives the roadmap. When it slows, deposition and etch follow. Inspection and metrology tools see fewer wafers.
Chart 1 shows the market’s verdict on this fragility over the past year.
Chart 1
Over the last 12 months, ASML has fallen 17.70%, dramatically underperforming peers and the broader technology sector. Applied Materials is down 7.86%, Lam Research up 10.00%, while KLA has gained 17.76%. The S&P Technology Select Sector Index rose 21.88% over the same period. In other words, all the stocks discussed in this article have underperformed the S&P.
This dispersion mirrors their risk profiles. ASML’s binary exposure to node progression leaves it most vulnerable to delays in scaling and China export headwinds. AMAT and Lam, though pressured, are cushioned by trailing-node spending. KLA stands out because inspection remains a non-negotiable requirement for all fabs, insulating it from leading-edge slowdowns.
The Larger Wafer-Front-End Picture and Extended Outlook
To see where this is heading, look at the larger WFE cycle. According to The Information Network’s report Global Semiconductor Equipment: Markets, Market Shares, Market Forecasts, WFE spending is projected to grow 2.5% in 2025, accelerate to 5.1% in 2026, then contract by 4.5% in 2027 before rebounding 8.0% in 2028. This is a mild upcycle, followed by a deeper mid-decade valley. For AMAT, Lam, and KLA, diversified portfolios allow them to ride out these oscillations. For ASML, the timing is more punishing. High‑NA EUV adoption would need to accelerate before 2027 to capture the current growth years. If it does not, the 2027 contraction will hit just as ASML is still ramping its most expensive platform ever. By the time WFE rebounds in 2028, the market may have adapted further to Low‑NA solutions, narrowing the urgency for High‑NA.
Chart 2 is my projected WFE spending growth trajectory that frames this dilemma.
Chart 2
This chart shows the mismatch between ASML’s R&D burn and the macro cycle. ASML is introducing the costliest lithography system in history into a tepid +2.5% market that will face contraction two years later. That leaves little margin for error. AMAT and Lam can fill revenue gaps with trailing-node shipments. KLA can sell inspection tools into any node. ASML has no such fallback.
Investor Takeaway
ASML is no longer the untouchable growth compounder it was during the 2018–2021 boom. It remains strategically essential, but strategically essential is not the same as financially predictable. With €10 billion sunk into High‑NA and no immediate revenue lift, ASML will face an extended shallow valley in returns. The China ceiling, TSMC’s Low‑NA pragmatism, and Intel’s contradictory outsourcing all point to a slower adoption curve. Meanwhile, AMAT, Lam, and KLA—while not immune—retain diversified demand streams. Their 3% drops on ASML’s earnings day were collateral damage. ASML’s 10.6% plunge was the direct hit. Until ASML proves High‑NA is more than a research tool, investors must see it for what it is today: a gamble tied to an unforgiving WFE cycle. Fouquet’s tariffs are just a distraction.