This illustrates the beneficiary structure of AI infrastructure, which is expanding to memory semiconductors, storage devices, server equipment, and power infrastructure. [Photo by Hanmi Ilbo Composite]

The theme for this week's market trend is 'From Memory to Infrastructure'.

This week's Stock Radar focuses not on making buy/sell decisions for specific stocks, but on understanding how the AI infrastructure cycle is expanding across which industry groups and supply chains.

Memory semiconductors were at the center of this week's stock market.

The AI semiconductor rally was not just about Nvidia. Related memory and storage stocks, including U.S. memory semiconductor company Micron Technology, and U.S. storage companies Western Digital, Seagate, and SanDisk, reflected expectations of AI infrastructure demand.

In the Korean market, SK Hynix and Samsung Electronics once again became the market benchmarks. While these companies do not dominate AI services, they are at the core of the supply chain that enables AI infrastructure.

As AI data centers increase, so does the demand for high-bandwidth memory, server DRAM, enterprise SSDs, and storage devices. The market has once again priced this connection this week.

This week, the Stock Radar asks one question: Do AI beneficiaries end with semiconductors, or do they expand to the entire infrastructure?

Three structures were confirmed in this week's market.

First, memory has been revalued as a bottleneck in the AI cycle.

As AI models grow larger, it's not just computational power that is needed. The ability to store, retrieve, and transfer data becomes equally important. This is why demand for high-bandwidth memory, as well as server DRAM, enterprise SSDs, and data center storage, is increasing together.

This week's rally in memory stocks reflects the market's interpretation that the AI cycle is expanding from computation to storage.

Second, AI infrastructure benefits are extending beyond semiconductors.

As data centers increase, power, cooling, networks, real estate, and private lending markets all move in tandem. The strong performance of U.S. network equipment company Cisco, driven by expectations of AI automation transformation and earnings forecasts, is also related to this trend.

AI is no longer just the name of a specific technology stock, but has become an investment language for the entire industrial infrastructure. Therefore, the market is not just looking at semiconductors, but is starting to re-examine companies related to power grids, equipment, networks, and data centers that are touched by AI.

Third, sharply rising stocks also faced profit-taking pressure simultaneously.

A good industry and a good price are different. The outlook that the AI cycle is valid and the judgment that short-term stock prices need a pause do not conflict.

The sharp declines and profit-taking seen in some semiconductor stocks this week are less an indication that AI demand has weakened, and more a signal that the market has begun to re-evaluate prices that have risen sharply in a short period.

This structure is more direct in the Korean market.

While Korean companies do not dominate the AI platforms, they are at the core of the supply chain that enables AI infrastructure. For this reason, large-cap Korean semiconductor stocks are evaluated as direct beneficiaries of the global AI investment cycle.

However, if a weak won, rising oil prices, and interest rate burdens occur simultaneously, stock performance and market valuations may move in different directions.

A weak won can be positive for export companies. However, if a weak won is combined with concerns about foreign capital outflow, it becomes a burden for the entire stock market.

This means it can be favorable for the profits of semiconductor companies, but it can act unfavorably for the discount rate of the Korean market.

What Stock Radar needs to look at this week is not the rise and fall of individual stocks, but which industry groups are being newly reflected in prices within the broader trend of AI infrastructure.

The points to watch next week are threefold.

First, we need to confirm whether the profit-taking in memory stocks will only result in a short-term adjustment.

Second, we need to see if AI infrastructure benefits are spreading to power, network, and equipment stocks.

Third, we need to examine whether the expansion of data center investment will remain a growth narrative or shift to a narrative of rising interest rates and debt burden.

The conclusion of this week's Stock Radar is as follows: While the AI cycle started with semiconductors, the market is now beginning to look at the prices and debts of all assets touched by AI.

※ This article is for reference analysis to understand market trends and industry changes, not an investment recommendation for specific stocks or financial products. The company names mentioned in the text are examples showing which industry signals the market reflected in prices this week, and do not represent buy/sell opinions or target price recommendations. Actual markets and stock prices may vary due to various factors such as interest rates, exchange rates, oil prices, policies, corporate earnings, and supply/demand changes, and the final investment decision and responsibility lie with the investor.