[New Samsung] ② Next M&A Targets: Infineon and NXP... Time to End the 'Lost 8 Years'

Fabless M&A is the Best Option but Not Easy
Memory Super Gap and Foundry Pursuit Require 'Selection and Focus'
AI, Bio, and Telecom M&A Difficult Due to 'Winner's Curse' Consideration

Samsung Electronics Chairman Lee Jae-yong's time embroiled in legal risks since the end of 2016 is referred to inside and outside the company as the "lost 8 years." During this period, competing companies have improved their business structures through mergers and acquisitions (M&A) worth tens of trillions of won in areas such as autonomous driving, artificial intelligence (AI) semiconductors, generative AI, semiconductor design, and NAND flash, but Samsung Electronics has not made any significant deals except for the acquisition of the $9 trillion-scale U.S. automotive electronics company Harman. The business community unanimously stated that after Chairman Lee's first trial acquittal, Samsung had the opportunity to pursue M&A in growth industries such as semiconductors, AI, and bio.

Among Samsung Electronics' M&A activities since the 2010s, notable cases include Medison (medical devices), SmartThings (Internet of Things), and Harman (automotive electronics). Since 2011, Samsung has successfully completed 28 major deals, but the only transaction exceeding the trillion-won scale was the acquisition of Harman (9.2 trillion won). This indicates that there were no deals considered to have secured Samsung's future growth engines.

This is distinctly different from the movements of competing companies. Intel acquired Mobileye, an Israeli startup specializing in autonomous driving, for $15.3 billion (about 20 trillion won), and AMD, the second-largest AI semiconductor company, purchased Xilinx (semiconductor design) for $50 billion (about 65 trillion won). Qualcomm acquired Nuvia (AI semiconductors) for $1.4 billion (about 2 trillion won), and Microsoft (MS) bought Sam Altman's OpenAI (generative AI) for $13 billion (about 17 trillion won). Domestically, SK Group acquired Hynix for about 3.4 trillion won. This deal is evaluated as having contributed to strengthening company momentum and improving business structure considering the transaction amount and outcomes.

Samsung Electronics Chairman Lee Jae-yong is entering the Seoul Gimpo Business Aviation Center (SGBAC) in Gangseo-gu, Seoul on the afternoon of the 6th to depart the country. [Image source=Yonhap News]

Experts say that among Samsung's growth businesses such as AI, semiconductors, bio, and 6G, semiconductor M&A is the most urgent. The ideal scenario is acquiring a fabless (semiconductor design specialist) company like ARM, which was mentioned after Chairman Lee's pardon and reinstatement in August 2022. Securing an outstanding design company would provide a foothold to manufacture various products such as automotive, server, AI, and mobile application processors (AP) depending on Samsung's sales network and market conditions. This would also give momentum to catch up with TSMC, the number one foundry (semiconductor contract manufacturing) company, in market share.

ARM was listed on the U.S. Nasdaq market in September last year, making it difficult for Samsung to acquire it. At that time, Samsung had secured around 100 trillion won in cash assets, so many evaluated that if the owner’s decision-making (approval) for spending tens of trillions of won had been made, it would have been a deal worth attempting. Even if competition authorities in various countries (including the Korea Fair Trade Commission) did not approve the deal, it would have served as an opportunity to imprint Samsung's investment sentiment in the market, according to the business community.

Realistically, the next targets for Lee Jae-yong's new Samsung M&A are automotive semiconductor companies such as Infineon (Germany) and NXP (Netherlands). If acquiring fabless companies is difficult, the strategy is to acquire top-tier semiconductor companies by segment and focus selectively. In particular, if Samsung had acquired NXP after Chairman Lee's pardon and reinstatement, it would have significantly contributed to expanding Samsung's system semiconductor (non-memory semiconductor) business portfolio, according to evaluations.

AI, bio, and telecommunications are also mentioned as long-term investment areas, but some opinions suggest that M&A is not necessarily required. It is considered too late to acquire large language model (LLM) companies with hundreds of billions of parameters, like OpenAI's ChatGPT. Rather, it is more realistic to focus on technology development strategies such as embedding Samsung Electronics' own LLM "Gauss" into on-device AI phones.

In the bio sector, since Samsung Biologics, an affiliate, has already become the world's number one company in contract development and manufacturing organization (CDMO), it is evaluated that entering the new drug business is more urgent. The dominant reaction is that acquiring new drug companies for cancer, heart disease, and rare disease treatments is too ambitious. Regarding 6G, since the global commercialization timeline is unclear, maintaining a business strategy centered on equipment contracts is considered the best option for now. The business community also analyzes that even the 5G telecommunications equipment orders, which Chairman Lee had worked hard on, have recently been cut off.

There are also concerns about the possibility of the "winner's curse" in M&A. Therefore, rather than pursuing M&A, it is suggested to clearly identify what the core semiconductor business is among memory such as DRAM and foundry, and aim to improve foundry yield (good product ratio), secure advanced processes, and expand customer base. The challenge for Lee Jae-yong's "New Samsung" is to strive to win the foundry three-way competition with TSMC and Intel.

Kim Kyung-joon, CEO Score's representative, said, "Samsung Electronics should observe how SK Hynix narrowed the market share gap in Samsung Electronics' DRAM from 20 percentage points to within 10 percentage points through the 'single-point breakthrough' of high-bandwidth memory (HBM)." He added, "It is necessary to clearly choose whether to focus on chasing TSMC's foundry or to decisively surpass SK Hynix and Micron in memory semiconductors and continue the 'super-gap' strategy."