South Korean President Lee Jae-myung, on June 29th, presided over a press conference for the "Three Super Projects," announcing that Samsung Electronics and SK Hynix will each construct two new memory chip fabrication plants in the southwestern region of the country.
The news triggered a significant rally in semiconductor equipment stocks. On Tuesday, ASML Holding NV (ASML) closed up 6.8% in Amsterdam, reaching a record high. In New York, Applied Materials (AMAT) closed approximately 4% higher, while KLA-Tencor (KLAC) gained 8%.
The rationale is straightforward: four new wafer fabs from South Korea's two memory giants imply substantial procurement demand for lithography machines, thin-film deposition equipment, and inspection systems. The primary suppliers for this equipment are precisely ASML Holding NV, Applied Materials, and KLA-Tencor.
Current Market Momentum and Outlook
This upward movement extends the robust performance of the semiconductor sector this year. The Philadelphia Semiconductor Index has nearly doubled in the first half and surged over 86% in the second quarter, marking its strongest quarterly performance on record. However, volatility within the sector remains high—the index fell 7.9% last week, its worst weekly performance since early April, before rebounding as investors returned to artificial intelligence infrastructure-related stocks.
Susquehanna analyst Mehdi Hosseini maintains a positive outlook on semiconductors, citing persistently strong industry demand. Other institutions forecast that global annual spending on wafer fabrication equipment will reach $250 billion by 2028.
Details of the Southwestern Korean Memory Cluster
The core of this investment plan is to establish a new memory chip cluster in southwestern South Korea, with a total investment of approximately 800 trillion won (about $518 billion). The goal is to double DRAM production capacity within five years. Samsung and SK Hynix will each build two fabs, with the government responsible for providing supporting infrastructure such as land, electricity, and water.
Simultaneously, Samsung separately announced a longer-term domestic investment plan on June 29th: from 2026 to 2040, it plans to invest a cumulative 2,450 trillion won domestically in South Korea. Of this, approximately 2,100 trillion won, or 76%, is earmarked for semiconductors.
A Goldman Sachs team led by analyst Giuni Lee provided a breakdown:
1,650 trillion won is allocated to existing fabs and projects under construction, including accelerating the completion of the Yongin Plant 6 from 2047 to 2040.
400 trillion won is designated for two new fabs in Gwangju, forming the core part of the southwestern cluster.
56 trillion won is for constructing a new HBM fab in the Chungcheong province.
Goldman Sachs estimates that, assuming Samsung's domestic capital expenditure and R&D account for about 80% of its consolidated total and projecting an approximate 6% annual growth rate, the cumulative domestic spending from 2026 to 2040 would be around 2,500 trillion won, aligning closely with the official announcement. Goldman Sachs therefore concludes that "this implied spending growth rate is not aggressive."
Capacity Doubling: A More Moderate Reality
While South Korean memory manufacturers have reiterated the goal of nearly doubling DRAM wafer capacity by 2030, a Bank of America Securities report issued on June 29th-30th, led by analyst Simon Woo, offers a dose of realism.
Doubling may sound aggressive, but it corresponds to a compound annual growth rate of only about 15%. More critically, when factoring in the closure of older fabs and the longer manufacturing cycles for next-generation memory chips, the annual growth rate of actual operational wafer capacity will be below 10%. The net wafer growth rate is projected to be in the single-digit percentage CAGR by 2030.
Bank of America Securities also notes that the new southwestern cluster is located far from the Seoul metropolitan area, requiring significantly larger infrastructure investment and posing greater construction challenges compared to existing bases like Pyeongtaek and Yongin. The firm draws a parallel to TSMC's dispersed layout strategy in Tainan, suggesting that such capacity expansion away from core regions requires a much longer lead time for preparation.
Considering the time needed for comprehensive infrastructure development (at least 5 years) plus fab shell construction and production ramp-up (an additional 3 to 4 years), Bank of America Securities estimates that the new cluster will not achieve meaningful production output for at least 8 to 10 years.
Upcoming Catalysts for the Equipment Sector
For the semiconductor equipment sector, the next critical juncture is the July earnings season. ASML Holding NV is scheduled to report results on July 15th, followed by TSMC on July 16th. Investors will closely scrutinize the latest capital expenditure guidance from these companies regarding new fab construction, as this will directly influence equipment demand expectations.
Comments