The Impact of Trump’s Second-Term Policies and the AI Investment Boom on the Semiconductor Market
Editor
A New Paradigm in the Semiconductor Market Triggered by Trump’s Second-Term Policies
As of the end of 2025, with President Donald Trump’s re-election confirmed, the U.S. semiconductor market is experiencing an unprecedented wave of change. President Trump’s public commitment to maintaining stock prices at record highs is interpreted by the industry not as mere political rhetoric but as a strong policy support for key technology sectors, including semiconductors. This statement is acting as a key driver for the stock price increase of major semiconductor companies, including California-based NVIDIA, and is directly impacting the restructuring of the global semiconductor supply chain.
The most notable aspect of the current U.S. stock market is the $600 billion annual capital investment by the Magnificent 7 (Apple, Microsoft, Google, Amazon, Meta, Tesla, NVIDIA). A significant portion of this investment is focused on AI infrastructure and data center construction, which implies an explosive increase in semiconductor demand. Washington-based Amazon Web Services (AWS) announced plans to invest $150 billion in data center expansion in 2025 alone, with approximately 40% expected to be allocated for purchasing high-performance AI chips. Cupertino, California-based Apple is also investing $20 billion annually in developing its own AI chips, intensifying competition with traditional semiconductor suppliers.
The Federal Reserve’s monetary policy is also having a complex impact on the semiconductor market. Despite inflation rates exceeding 3%, the Fed’s decision to lower the benchmark interest rate is interpreted as a signal of prioritizing economic growth. This accommodative monetary policy lowers the financing costs for semiconductor companies, promoting large-scale investments, while simultaneously presenting a dual challenge of rising manufacturing costs due to inflationary pressures. Particularly, memory semiconductor companies like South Korea’s Samsung Electronics and SK Hynix are experiencing improved export competitiveness due to a weaker dollar, but are also facing cost pressures from rising raw material prices.
The most noteworthy change is that the global AI infrastructure investment scale is reaching $1 trillion annually. This figure represents a 150% increase compared to 2023, significantly surpassing the total revenue scale of the semiconductor industry, which stands at $574 billion. Most of this investment is concentrated on GPUs, HBM (High Bandwidth Memory), and AI-specific processors, with demand for cutting-edge AI chips like NVIDIA’s H100 and next-generation B200 series far exceeding supply. Taiwan Semiconductor Manufacturing Company (TSMC), headquartered in Hsinchu, Taiwan, announced plans to increase its capital investment by 35% year-on-year to $36 billion in 2025 to respond to this surge in demand.
NVIDIA’s Dominance and the Restructuring of the Semiconductor Ecosystem
The fact that NVIDIA’s market capitalization surpasses the stock markets of most countries worldwide, except the top five, starkly illustrates the concentration in the current AI semiconductor market. As of November 2025, NVIDIA’s market capitalization exceeds $3.2 trillion, more than 50% higher than the entire market capitalization of South Korea’s KOSPI, which stands at $2.1 trillion. Even more astonishing is that NVIDIA’s market capitalization alone is equivalent to 60% of the entire market capitalization of the Tokyo Stock Exchange in Japan. This phenomenon reflects NVIDIA’s monopolistic position in the GPU market, the core infrastructure of the AI revolution.
However, cracks are beginning to appear in this dominant system. Santa Clara, California-based AMD (Advanced Micro Devices) is directly competing with NVIDIA in the data center AI market with its MI300X series, recording a 122% year-on-year increase in data center revenue to $6.9 billion in the third quarter of 2025. Intel is also attempting to expand its market share through its Gaudi3 AI accelerator, particularly targeting companies in China that find it difficult to purchase NVIDIA products due to U.S. export restrictions. Meanwhile, Google’s TPU (Tensor Processing Unit) and Amazon’s Inferentia chips, initially developed for internal use, are now being offered as services to external customers, posing a new challenge to the traditional chip sales model.
In the memory semiconductor market, South Korean companies are emerging as the biggest beneficiaries of the AI boom. Samsung Electronics announced plans to increase HBM3E production by 300% compared to 2024, while SK Hynix is leading the industry with over 50% market share in the HBM market. Notably, SK Hynix’s HBM3E products are exclusively supplied to NVIDIA’s next-generation GPUs, the B200 and H200, with HBM revenue in the fourth quarter of 2025 estimated to have increased by 400% year-on-year. These achievements indicate the exponential growth in memory bandwidth and capacity required for AI training and inference processes.
The announcement that the Federal Reserve’s quantitative tightening will end in two days is interpreted as a signal of additional liquidity supply to the semiconductor industry. With the conclusion of the quantitative tightening policy, which involved the Fed selling $95 billion worth of treasuries and MBS (mortgage-backed securities) monthly over the past two years, more funds are expected to be supplied to the market. This is analyzed to facilitate the financing needed for large-scale capital investments and research and development by semiconductor companies. Indeed, TSMC announced plans to invest an additional $25 billion in the construction of a fab in Arizona following the announcement of the end of quantitative tightening, aligning with the Trump administration’s policy direction to expand semiconductor production capacity within the U.S.
The Long-Term Impact of Fiscal Policy Changes on the Semiconductor Industry
The current situation where U.S. fiscal deficit spending exceeds 6% of GDP is having a complex impact on the semiconductor industry. This expansionary fiscal policy is acting as a factor promoting semiconductor demand in the short term. Particularly, government infrastructure investments and defense spending provide a stable foundation for semiconductor demand, with the $52.7 billion support for semiconductor manufacturing through the CHIPS Act continuing. Intel is utilizing this support to construct a new $20 billion fab in Ohio, aiming for operation in 2026. However, in the long term, the expansion of the fiscal deficit may exacerbate inflationary pressures, potentially leading to increased semiconductor manufacturing costs.
The forecast that corporate stock buybacks will reach a record $1.2 trillion in 2026 is bringing changes to the shareholder return policies of semiconductor companies. NVIDIA approved a $25 billion stock buyback program in 2025, the largest in the company’s history. Microsoft also announced a $60 billion stock buyback, demonstrating confidence in the growth of the cloud services sector closely related to semiconductors. These large-scale stock buybacks exert upward pressure on stock prices, while raising concerns that using a significant portion of cash for shareholder returns may limit the capacity for new investments.
President Trump’s proposal to abolish income tax is expected to positively impact the semiconductor industry’s ability to secure high-level talent. The U.S. semiconductor industry is currently facing a severe talent shortage, particularly in AI chip design and advanced manufacturing processes, where it is extremely difficult to find experienced engineers. If income tax abolition is realized, the influx of overseas talent into the U.S. is expected to accelerate, potentially enhancing the innovation capabilities of semiconductor companies. In fact, major semiconductor companies like NVIDIA, AMD, and Intel are already hiring a large number of high-level talents from India, China, and South Korea, and this trend is expected to accelerate further with additional tax benefits.
The promise of a $2,000 stimulus check in 2026 is analyzed to have an indirect impact on the semiconductor market through increased consumer electronics demand. Considering the experience of the pandemic-era stimulus policies leading to a surge in demand for smartphones, laptops, and game consoles, resulting in a semiconductor supply shortage, the new stimulus package is likely to have a similar effect. Qualcomm has already announced plans to increase 5G smartphone chip production by 25% by the first quarter of 2026, and MediaTek is also expanding its supply of chipsets for mid-range smartphones. These preparations are interpreted as being in anticipation of increased consumer spending following the implementation of the stimulus package.
The current semiconductor market situation is the result of the complex interplay of the AI revolution, geopolitical competition, and the U.S.’s expansionary economic policies. In the short term, the explosive growth of the AI chip market centered around NVIDIA and the benefits to South Korean memory companies are expected to continue. However, in the medium to long term, intensified competition, geopolitical risks, and macroeconomic imbalances are likely to emerge as new challenges. Particularly, as China’s efforts to build its own semiconductor ecosystem and Europe’s semiconductor sovereignty policies become more pronounced, the restructuring of the global semiconductor supply chain seems inevitable. In this environment, companies face the new challenge of simultaneously strengthening their technological innovation and geopolitical risk management capabilities.
