Amidst the competition for AI supremacy, there is a hidden group that holds the potential to emerge as the ultimate victor. Referred to as the 'undiscovered' group, these entities operate as crucial intermediaries within the semiconductor sector and are poised to benefit from multiple favorable factors.

 




Stock quotes mentioned in this article include NVDA (Nvidia), AMD (Advanced Micro Devices), AMZN (Amazon), GOOGL (Alphabet), AAPL (Apple), AMAT (Applied Materials), KLAC (KLA Corp.), LRCX (Lam Research), ASML Holding (ASML), INTC (Intel), TSM (Taiwan Semiconductor), and TER (Teradyne).


The semiconductor capital equipment sector, often likened to the "arms dealer" of the semiconductor industry, plays a crucial role in supplying the necessary capacity for the ongoing battle for AI supremacy.


Nvidia (NVDA) holds an early advantage in AI chips and requires significant capacity expansion to meet the surging demand. Meanwhile, Advanced Micro Devices (AMD) plans to introduce its M1300X chip later this year, and tech giants like Amazon (AMZN), Alphabet (GOOGL), Apple (AAPL), and others are developing their own AI-related silicon.


Regardless of the outcome of the AI supremacy race, semiconductor capital equipment companies such as Applied Materials (AMAT), KLA Corp. (KLAC), Lam Research (LRCX), and ASML Holding (ASML) stand to benefit.


One positive aspect for the semiconductor capital equipment sector is the increasing government incentives that attract substantial investments in new chip plants. Initiatives like the CHIPS Act, combined with worldwide incentives, are expected to total around $66 billion annually over the next five years, encouraging domestic manufacturing capacity. Intel (INTC) recently announced multi-billion dollar chip-making plants in Israel, Poland, and Germany, taking advantage of incentives totaling $12 billion.


Intel's expansion of foundry capacity to compete with Taiwan Semiconductor (TSM) and Samsung not only offers diversification in chip supply but also addresses concerns about overreliance on Asian manufacturing. As supply chains shift away from China and concerns about chip manufacturing concentration persist in Taiwan, Intel's foray into the foundry industry will be closely monitored.


The semiconductor capital equipment stocks are experiencing tailwinds from new plant developments, alongside the growing demand for AI-related orders.


Jefferies identifies the semi-cap equipment sector, including Applied Materials, Lam Research, KLA Corp., and Teradyne, as undiscovered plays in the AI space. They highlight the need for larger parallel processing chips in AI, which drives demand for semi-cap equipment. Advancements in chiplet architectures further require advanced packaging and testing equipment.


It should be noted that while the "undiscovered" thesis is still in the discovery process, the progress in AI will continue to drive increased capital expenditure by chipmakers.


Nvidia is reportedly seeking up to 50% more supply from Taiwan Semiconductor for 2024, with its wafer use already increasing by nearly 50% since the beginning of 2023. Wall Street closely monitors Taiwan Semi's capex needs, expected to be between $32 billion and $36 billion for 2023, as the demand for AI is expected to have a ripple effect throughout the semi-cap equipment sector.


The sector has witnessed significant profitability growth over the past decade, aided by improved margins resulting from consolidation. Valuations of the group remain reasonable, with P/E multiples typically ranging between 16x and 18x. Additionally, companies in this sector return a significant portion of cash flow to investors through buybacks and dividends.


On the bearish side, concerns arise from the aggregate expansion of new factories and capacity at a time when wafer fab equipment/sales are at a 15-year high. There is a risk of industry over-investment leading to a significant downturn. However, the semiconductor industry seems to be emerging from a cyclical downturn as excess inventory is currently being worked down.



Comments