You know how stocks become more expensive when investors worry about even the smallest change in earnings expectations. This can be seen in the sharp movements of stocks related to AI in recent weeks, especially the best options in the sector such as NVIDIA, ASML, ARM, and Asian and US chip manufacturers.
Take ASML for example. The logic behind these moves appeared to be the declining expectations for the growth of artificial intelligence chips, as device orders for the world’s largest chip equipment manufacturer did not exceed market expectations .
Certainly, the conditions are ripe for a correction, as prices in the AI stock market have risen strongly and very quickly. This boom has marked disproportionate gains even for stocks that don’t appear to be the biggest beneficiaries of AI-driven growth. At the same time, the region remains highly exposed to geopolitical risks. China is an important market for most chip-related companies, including Nvidia and ASML, and in the case of the latter, China is its largest market and was making up about half of their system sales in the first quarter.
However, the recent sharp volatility in AI stocks illustrates the market’s true sensitivity to relatively mild cyclical movements more than anything else. In fact, the AI chip market could continue to grow rapidly despite the apparent weakness in the broader semiconductor sector, as chips made for AI applications, despite notable growth were sold last year, still a very small percentage of global supply. . Broadband chips are an example of this, and are an integral part of all AI chips. A single AI chip can only be run with six such advanced memory chips, including NVIDIA chips.
These chips currently account for about 1% of the total memory chip market, measured by sales volume, and even when measured at today’s growth rates, artificial intelligence chips cannot to compensate for the occasional decline in the market in general, which is still dominated by traditional memory, storage. , and processing chips used in smartphones, cars, and electronic devices.
In this context, companies are currently using their surplus stocks, which are the result of overproduction by manufacturers, and the accumulation of machinery manufacturing companies due to the shortage that hit the market during the pandemic. Higher-than-normal inventories have dampened end-market demand, so chip manufacturers have not felt the urgent need to increase production capacity or place new orders for chip manufacturing equipment. in recent years. Memory chip prices have fallen more than 50% since 2021, and spot price reductions continue this month. Demand recovery remains particularly slow among car and smartphone manufacturers, two of the biggest chip buyers.
Additionally, seasonal variations in ASML orders were generally not a concern. The company sells about 100 new lithography systems every quarter. Three companies buy most of the company’s advanced equipment, making quarterly readings subject to change. Recording readings at lower than expected levels in a quarter does not necessarily reflect the results that will be announced for the rest of the year.
In contrast, a prolonged decline in chip manufacturers’ orders for equipment compared to estimates likely means they are more likely to place larger orders in the coming quarters. If we look at the rest of the world, we find that countries around the world are working on building more than 70 new chip manufacturing projects. TSMC and Samsung are expected to begin mass production of the next generation of chips using 2-nanometer technology next year. In this regard, companies must submit orders for new equipment this year, produced exclusively by ASML, for the new production lines.
The latest TSMC products help prove this. Although he lowered his chip market growth expectations excluding memory chips for this year, he kept his plans for capital spending for the year unchanged, to remain between $28 billion and $32 billion. The company is currently working on building new factories in countries including the United States, Japan and Germany, as well as plans to build factories to start making chips with real technology. advanced 1.6 nanometer in 2026.
It’s worth noting that the main drivers for the rise in AI stocks over the past two years had little to do with the cyclicality of the traditional chip and equipment sectors, but a lot to do with the transformative potential of the AI industry. For those anticipating the peak of AI growth, they should focus on the difference between the severity of valuations and the expected slower pace of wider AI adoption, not e the normal distribution factors.
2024-05-05 22:02:28
#stock #correction #overdone