ASML, the lithography giant with the reputation of "the pinnacle of human technology", had a rare disappointing performance, causing a global semiconductor stock plunge. ASML's weak performance undoubtedly poured cold water on the recent rebound of semiconductor stocks from the "summer selling wave". Incomplete statistics show that since the opening of the US stock market on Tuesday, until the close of the Japanese and South Korean stock markets on Wednesday, the total market capitalization evaporated by over $420 billion, including Nvidia, AMD, SK Hynix, and Samsung among other global semiconductor stocks.

The process of ASML's performance announcement can be described as full of drama. ASML was originally scheduled to release its third-quarter financial report on Wednesday, but due to "technical issues", the financial report was unexpectedly posted on the company's official website in the early trading hours of US stock market on Tuesday, creating a "fracture-level" embarrassment. For such a technological giant that controls the core lifeline of global chip production capacity, such a significant error is a rare occurrence of historic proportions.

As a result, ASML's exceptionally poor performance has surprised global investors: although overall revenue exceeded expectations, the order volume was only half of the market's expectations, and next year's sales target and gross margin guidance were unexpectedly reduced. In this "prematurely leaked" third-quarter financial report, ASML stated that the total order volume in Q3 was only 2.6 billion euros, nearly half of the market's general expectation of 5.4 billion euros. At the same time, the company expects the total net sales to increase to between 30 billion and 35 billion euros by 2025. However, compared to this, the previous guidance given by the company for the previous quarter was 30-40 billion euros. This latest forecast figure is only in the lower half of the guidance provided at the Investor Day in 2022, so the sales outlook has been significantly reduced by ASML management.

After the financial report was released, ASML's ADR in the US stock market fell by more than 17% at one point, eventually closing down by 16.26% at $730.43. The company's stock price on the Amsterdam stock market fell by 16%, marking the largest single-day decline since 1998, with multiple trading halts during the day. Following ASML's sharp decline, the title of the "most valuable tech company in Europe" in the European stock market changed hands, with the long-held title being transferred from the lithography giant ASML to the German software leader SAP, which has greatly benefited from the AI ​​boom.

In addition to the disappointing performance, there are reports that the US government is considering restricting American companies from selling the most advanced artificial intelligence chips to certain Middle Eastern countries, including the high-demand flagship data center AI chips from Nvidia and AMD. This news further drove the sharp decline in ASML's stock price, as the demand for AI chips from giants like Nvidia may decrease, potentially prompting TSMC to reduce the production capacity of AI chips for Nvidia, leading to a downward revision of ASML's performance expectations as a result of this logic.

However, from a rational investment perspective, ASML's performance, which has hit the global chip stock price hard, does not mean that the global frenzy of AI deployment is dissipating or cooling down. However, this disastrous financial report has indeed exposed the latest dynamics of the global chip industry, which is that the AI ​​boom is still ongoing. Especially, all types of AI chips focused on B-end data centers are still in high demand. However, areas not related to AI are still facing weak demand, even a significant decline in demand.

ASML CFO Roger Dassen has supported this market view in the performance statement. The ASML executive stated that the demand for AI-related chips is indeed increasing rapidly. However, the recovery process in other parts of the semiconductor market is weaker than expected, leading to some logic chip manufacturers postponing lithography machine orders.

Before ASML's financial debacle, there has long been an optimistic outlook in the stock market, that a new round of the entire chip industry's 'prosperity cycle' has quietly started in 2024. However, ASML's latest performance has cast doubt on this bullish view, triggering a sell-off of chip stocks in the market. In other words, the global AI frenzy, laid out by global enterprises and even global government institutions, has not waned. However, this AI frenzy has not yet driven the entire chip industry, including long-term demand in weak areas such as industrial chips, electric vehicle chips, power chips, components, and most consumer core chips including smartphones, into a 'prosperous development period'.

After the's ASML financial failure, some analysts even shout out, even if ASML's performance is so bad, it actually benefits NVIDIA and other 'AI shovelers' like global AI chip demand are still strong. From the stock price trend, ASML, which holds the lifeline of chip capacity, has significantly underperformed NVIDIA. Unconsciously, the market has given a clear answer in terms of who is the biggest winner of chip stocks with real silver and gold.

ASML revaluing expectations may mean a slight excess of chip production capacity, not the end of the chip industry.

Regarding ASML's weak performance report, some chip industry analysts believe that the bleak performance outlook may reflect some excess capacity in chip factories. Large chip foundries including TSMC, Intel, and Samsung have accumulated a large number of expensive ASML EUV tools during the COVID-19 pandemic and these foundry giants are good at transforming and upgrading these tools to produce more high-end AI chips.

Currently, apart from AI chips, the inventory of other types of chips remains high. At the same time, after years of process exploration, chip manufacturers such as TSMC have significantly improved the efficiency of using ASML lithography machines, which means they can produce more chips and are not in a hurry to order more lithography equipment.

For example, due to the lack of significant increase in demand for smart phones and PCs, Taiwan Semiconductor, known as the 'chip manufacturing king', may refurbish and upgrade EUV for some 5nm and below smart phone and PC chips to boost demand for the 'never-ending' Nvidia H100/H200 AI GPU, the latest Blackwell architecture AI GPU, and the AMD MI300 series GPU, rather than continuing to invest heavily in expensive EUV.

Of course, during the extremely hot demand period for Nvidia AI GPUs from 2023 to 2024, ASML's financial reports for multiple quarters show that TSMC continuously purchases new EUV machines, but the scale of purchases is not as large as some analysts predicted. In addition, the pace of purchases may be stagnant in the third quarter, fully highlighting that as the demand for non-AI chips continues to decline, TSMC and other foundries will transform and upgrade excessive lithography equipment accumulated since the pandemic to manufacture high-end AI chips, reflecting to a large extent that the AI frenzy has not driven the entire chip industry into a growth cycle, and chip factories' capacity is in a slight surplus state.

Therefore, TSMC and other chip foundries can only order new EUV or DUV lithography machines when factory orders are full. Some analysts also indicate that ASML's latest forecasts are lagging indicators of the poor performance of these chip factories in recent months, revealing the latest situation of sluggish demand in the chip market for non-AI applications.

"Intel, TSMC, and Samsung may be retracting orders for lithography machines from ASML, as they have realized that the capacity is very adequate." Dan Hutcheson, Vice President of the technology industry-focused analysis firm TechInsights, stated in a recent interview.

"This year, chip factories have an approximately 81% capacity utilization rate, but manufacturers tend to buy new lithography tools when it reaches around 90%. In addition, Intel has slowed down the pace of expansion of its U.S. factories, and Samsung and TSMC have also taken a cautious stance on expanding chip production capacity in the near future." Hutcheson stated.

Handel Jones, CEO of an international business strategy company tracking the chip manufacturing industry, stated that some chip foundries have reduced the steps using ASML's flagship lithography machines, with some reducing them by almost one-third. Using Samsung as an example, he mentioned that Samsung may be able to use more advanced chip etching technology in the future, reducing the steps involving ASML's flagship lithography equipment from five or six to just one or two steps. He explained that if successful, Samsung may face a significant excess of chip manufacturing capacity on these EUV lithography machines.

However, Jones, who has been tracking the dynamics of the chip industry for many years, pointed out that he has not changed his forecast for the entire chip industry. With the thorough innovation of various industries' productivity after the update and iteration of large AI models, the demand for AI chips and AI dedicated storage chips will thrive in the future, inevitably driving the entire chip industry, including industrial simulation chips, electric vehicle chips, and consumer electronics SoC chips, into a prosperous cycle. "This is just a short-term fluctuation in the chip industry. In the long run, everything will improve."

The AI frenzy is still in full swing, but the entire chip industry has not fully benefited from it.

The performance results of ASML indicate a significant differentiation in the fate of global chip companies: ChatGPT and other artificial intelligence applications like Sora that can handle massive parallel computing patterns and high computing density matrix operations on data center server-side AI chips have seen a surge in demand for NVIDIA AI GPUs, masking the extremely low demand in other sub-sectors of the industry.

ASML did not elaborate on why its third-quarter orders were less than half of the analysts' forecasts, simply stating that some customers postponed factory construction. Analysts speculate that the "customers" referred to by ASML should be Intel and TSMC - the two chip manufacturers have previously indicated a slowdown in chip factory construction globally. Faced with shrinking chip product sales and increasing net losses in chip foundry business, Intel is significantly slowing down chip factory construction processes, postponing plans to build new factories in Germany and Poland last month.

Jefferies Financial analyst Janardan Menon from Wall Street stated in a report on Wednesday: "ASML's financial report shows that although demand for chips related to artificial intelligence remains very strong, recovery in other areas is exceptionally lagging, and this trend may continue until 2025."

In a recent report, analysts from another major institution Bernstein expressed concerns about the longer time needed for the recovery of end demand, leading to significant delays in chip capacity expansion, which is what we will be facing by 2025. "It seems we may need to wait patiently until the cyclical recovery of the chip industry becomes clearer."

The latest semiconductor industry outlook data released by the World Semiconductor Trade Statistics Organization (WSTS) shows that AI-driven storage chips, GPUs, CPUs, and other logic chips are the main driving force behind the recovery of the entire chip industry. In contrast, WSTS' expectations for the analog and microchip market size covering electric vehicles, industrial sectors, IoT devices, as well as analog and microchips for more widely consumed electronic products such as PS5, Switch, appear very bleak, with this year's market size forecast for the analog chip sector even negative growth, with hopes of achieving a weak recovery process next year.

WSTS states that the revised expectations for 2024 reflect strong performance over the past two quarters, especially in the computing terminal market. After a significant contraction in the market in 2023, WSTS expects that in 2024, there will mainly be two core chip product categories driving double-digit sales growth in 2024, namely a 10.7% increase in total sales of logic chips, including the in-demand NVIDIA AI GPU and AI ASICs, and a substantial 76.8% increase expected in the storage chip category that best reflects chip cycles for 2024 - WSTS predicts strong demand for storage chips focused on HBM used in AI training/inference fields and enterprise-level DRAM and NAND storage systems crucial for efficient operation of AI data centers.

The situation of AI-driven storage and logic chip demand is very evident from the inventory and export scale of chips in South Korea, which is the home of the world's largest two memory chip manufacturers - SK Hynix and Samsung.

Data released by the South Korean government shows that although the growth rate has slowed down, semiconductor exports in September still increased significantly by 37% year-on-year, marking the 11th consecutive month of growth. This growth was slightly weaker than the 38.8% increase in August. Within the continuous growth of chip export data, one-third of the growth was contributed by HBM storage systems. The HBM storage system, along with the core hardware provided by AI chip leader NVIDIA - H100/H200/GB200 AI GPU, is crucial for driving heavy-duty artificial intelligence applications such as ChatGPT and Sora. The stronger the demand for HBM, the more intense the demand for AI chips.

US banking giant Bank of America recently released a research report stating that the global AI boom is still in its early stages, similar to the development path of the internet in the 1990s. It can be likened to the '1996 moment' of the flourishing development of the internet. According to the Bank of America analysis team, the AI boom is still in a very early stage.

"Investors may be underestimating the long-term impact of this technology while overestimating its short-term potential, but this is also a typical feature of technological prosperity. It is expected that in the coming years, capital expenditures related to artificial intelligence may exceed $1 trillion, but compared to the internet age, the development of artificial intelligence is just getting started," Bank of America pointed out in the report.