The global artificial intelligence (AI) and semiconductor markets are going through unprecedented vo..

The global artificial intelligence (AI) and semiconductor markets are going through unprecedented volatility. Meanwhile, domestic asset managers' AI active exchange-traded fund (ETF) strategies have also been divided into "memory-intensive" and "big tech decentralized." Despite being the same "U.S. AI Active," the gap in returns has widened to 13 percentage points in a month, which seems to complicate investor counting methods.

Recently, global semiconductors and AI-themed products are diversifying in the domestic ETF market. Among them, four active ETFs are typically listed, focusing only on stocks listed on the U.S. stock market. According to the Korea Exchange on the 7th, the yields of these four ETFs for the past month as of the previous day differed according to the proportion of memory companies being incorporated.

ACE U.S. AI Tech Core Industry Active (-10.09%) by Korea Investment Trust Management and RISE U.S. AI Tech Active (-11.94%) by KB Asset Management received relatively good report cards as a result of diversifying into Big Tech. On the other hand, Time Folio Asset Management's "TIME Global AI AI Active" fell to -14.72%, and Mirae Asset Asset Management's "TIGER Global AI Active" fell to -22.70%.

ACE U.S. AI Tech's core industrial activities include NVIDIA, Palantir, Meta, and TSMC at around 9%, respectively. The main difference is that it includes AI infrastructure companies evenly. It has incorporated data center cooling company Vertiv Holdings, small module nuclear power plant (SMR) company Oclo, and uranium mining company Cameco. Analysts say that the recent surge in stock prices has been effective in defending performance.

RISE USA AI Tech Active's portfolio is most focused on Magnificent 7 (M7). Big technologies such as Nvidia, Microsoft, Amazon, Alphabet, and Apple are evenly included at around 9%. The total number of stocks is also 40 stocks, taking a diversified investment strategy.

TIME Global AI Artificial Intelligence Active and TIGER Global AI Active, which have a high proportion of memory, have something in common that their returns have been sluggish recently. However, there was a difference in major strategic events. TIME Global AI Artificial Intelligence Active chose Kioxia and TIGER Global AI Active selected AMD as the highest weight category. In addition, it includes representative stocks of memory and storage devices such as Sandisk, Western Digital, Intel, and Seagate. In particular, in the case of TIME Global AI Artificial Intelligence Active, the aftermath of the recent month's plunge in memory stocks was noticeable compared to the 48% return on the basis of the past six months.

Recently, memory companies have repeatedly seen their stock prices plunge even after the announcement of good performance. SanDisk, a representative storage unit, plunged 6% in after-hours trading after the company reported earnings on the 5th (local time). This is because despite its solid performance, it did not meet high market expectations. Despite a 372% year-on-year surge in sales in the fourth quarter of 2026, SanDisk's sales grew as its sales forecast for the next quarter fell short of expectations. This led to worsening investment sentiment in the global storage device industry as well as domestic semiconductor stocks such as SK Hynix (-9.7%) and Samsung Electronics (-6.1%).

However, the industry is widely criticized for excessive memory price concerns. According to market research firm TrendForce, DRAM and NAND prices have only slowed down, but are still on the rise. Wall Street diagnosed that there is a "decoupling (asynchronization)" in which real memory prices rise but corporate stock prices go down. The U.S. investment magazine Barance said, "It is not because of the damage to the actual business, but because of the burden of valuation," adding, "The memory for AI servers is an opportunity for long-term investors to buy on dips as supply continues to be in short supply."

While memory stocks faltered, big tech companies continued to rise. CNBC, a U.S. economic media, reported on the 31st of last month that "Big Tech's market capitalization has increased by $1.5 trillion due to the expansion of AI facility investment and strong performance of three cloud companies, Microsoft, Amazon, and Alphabet." This is the result of confirming solid cloud growth.

However, there are still voices of concern about the expansion and sustainability of AI investment. Alphabet posted its first quarterly free cash flow (FCF) deficit in the second quarter due to large-scale AI investment, and Meta's market capitalization of $85 billion evaporated during the same period.

Nevertheless, the stock market saw that Big Tech's virtuous investment cycle structure is valid. Bloomberg analyzed that the free cash flow of the five major hyperscalers (Microsoft, Meta, Alphabet, Amazon, and Oracle) will be in the red until 2027, but will turn positive after 2028.

Yang Yang-woo, a researcher at Samsung Securities, said, "Big Tech's cash flow seems sluggish because funds are moving to semiconductor and infrastructure companies through facility investment," adding, "The free cash flow of the entire AI value chain will continue to increase."

Experts believe that semiconductor volatility will continue until the winner of the AI era is decided. In particular, there are observations that volatility will increase in the second half of the year due to interest rate uncertainties, macroeconomic environment, and doubts surrounding the semiconductor and big tech industries. "As large-scale initial public offerings (IPOs) remain in the U.S. stock market this year, instability in supply and demand will also affect semiconductor stocks," said Lee Kyung-joon, head of Kiwoom Investment Asset Management. This is why ETFs are drawing more attention in semiconductor and AI investments.

Furthermore, the benefits of AI semiconductors are expected to spread from simple graphics processing units (GPUs) to central processing units (CPUs) and big tech's own on-demand semiconductors (ASICs). In the short term, it is advised that it is advantageous to prevent downside risks by utilizing fabless and customized chip ETFs that benefit from Big Tech's own chips and ETFs that are distributed to AI infrastructure such as power and nuclear power plants.

"The recent volatility of the stock market is not limited to the Korean market," said Park Seung-jin, head of Hana Securities' overseas stock analysis division. "Active ETFs can be more advantageous than buying individual stocks as fund managers preemptively adjust their share and weight depending on the issues of individual stocks." However, he added that the management company should check the asset composition details (PDF) disclosed every day. "If you are a long-term investor, you can also consider a strategy to buy a low-priced 'active ETF centered on memory and storage devices', which has plummeted," he advised.