Betting on AI! Hedge fund mogul Dan Loeb is eyeing these two stocks and continues to hold Taiwan Semiconductor and NVIDIA

Hedge fund Third Point founder Daniel Loeb is doubling down on the AI sector, expanding his portfolio to include what he believes are undervalued international semiconductor companies.

On November 3rd, Third Point released its latest third-quarter investor letter, disclosing its positions in South Korean memory chip manufacturer SK Hynix and Japanese semiconductor production equipment manufacturer Ebara.

Loeb stated in the letter that both companies are leaders in their respective industries and are significant beneficiaries of AI infrastructure development, with valuations that are "absolutely reasonable" and even "significantly discounted" compared to their American counterparts.

Although Loeb acknowledged that his fund's performance in certain areas has not met expectations recently, he expressed an overall optimistic outlook on the market. He believes that ongoing AI investments and the potential interest rate cuts by the Federal Reserve indicate a favorable market environment is likely to continue.

Computing Power Remains King, Continuing to Favor NVIDIA

Despite expanding into new investment territories, Loeb's conviction in core assets related to AI computing power remains strong. He emphasized:

We still live in a world constrained by computing power.

He pointed out that this theme has benefited his existing investments in Taiwan Semiconductor and NVIDIA, both of which are indispensable companies in AI infrastructure development.

Loeb's team has conducted in-depth research on the development of AI technology. He countered previous concerns about "peak AI computing demand" triggered by breakthroughs in training efficiency from models like DeepSeek. He wrote:

Reality has proven to be entirely different. The new AI capabilities and architectures, especially inference-based models, are several orders of magnitude more compute-intensive than their predecessors.

Loeb believes this not only offsets efficiency gains but also leads to a significant acceleration in AI computing demand.

Loeb concluded that AI computing has expanded from a singular pre-training expansion law to three dimensions: pre-training, post-training, and inference.

Beware of Risks, Closely Monitor AI Infrastructure Investment and Credit Markets

While optimistic about the AI outlook, Loeb also remains vigilant about potential risks.

In the letter, he stated that his team is "closely monitoring any risks in the ecosystem that could signal a potential pullback." He warned:

One area we are closely watching is AI infrastructure, particularly the unprecedented scale of investment in power and data centers.

He noted that the scale of capital investment is "several times" that of the entire high-yield bond and leveraged loan market. While he acknowledged that "its impact (and opportunities) will be enormous," he also cautioned that "demand and supply are unlikely to grow in sync and smoothly."

Additionally, the fund is actively seeking and seizing multiple opportunities in the credit market.

Loeb mentioned that the corporate credit investment return in the third quarter reached 4.0% (net return 3.7%), with a year-to-date cumulative return of 7.2% (net return 5.4%).

Moreover, specific events this year have created several distressed asset trading opportunities, such as subprime auto loan originator Tricolor and subprime auto parts manufacturer First Brands The former has seen its bonds sold off due to credit issues, and the fund is assessing its capital structure and has made a small investment. The latter experienced a sharp decline in the price of senior loans after bankruptcy, and the fund is analyzing the impact through its collaborative team and profiting from the price fluctuations of collateralized loan obligations (CLO) tranches