Major memory semiconductor companies and leading U.S. tech firms are set to announce their earnings this week. While demand for AI infrastructure is translating into actual semiconductor orders and rising prices, questions are growing about whether these companies can generate sufficient profits to recoup their massive investment costs. This earnings season is expected to serve as a litmus test for the growth and sustainability of the AI industry.
Starting on the 28th, SK Hynix will kick off the earnings announcements, followed by Samsung Electronics on the 29th, Japan's Kioxia on the 30th, and China's ChangXin Memory Technologies (CXMT) on the 31st.
The combined market capitalization of these four major memory semiconductor companies exceeds $2 trillion. As memory semiconductors are key components in AI data centers, their performance will indicate how much AI infrastructure investment is translating into actual semiconductor demand.
The outlook for the memory giants is positive. Global tech firms are expanding investments in AI data centers and computing infrastructure, leading to increased semiconductor demand and rising product prices due to supply shortages.
Samsung Electronics previously reported an estimated operating profit of 107.4 trillion won for the third quarter. SK Hynix is also projected to exceed 75 trillion won in operating profit for the same period.
Kioxia's performance is expected to improve significantly. According to Bloomberg, analysts predict that Kioxia's operating profit for the September quarter will increase more than 20-fold compared to the same period last year.
CXMT also forecasted on the 17th that its third-quarter revenue will reach between 114.7 billion and 119.7 billion yuan, with net profit expected to be between 53.4 billion and 54.9 billion yuan. If these estimates hold, it would mark the highest third-quarter performance in the company's history. Cumulatively, net profit for the first three quarters is expected to reach up to 270 billion yuan, more than eight times higher than the same period last year.
However, strong earnings from memory semiconductor companies alone do not prove the profitability of AI investments. Even if semiconductor prices and order volumes remain robust, it remains uncertain whether big tech firms can recover the substantial funds they have invested in AI infrastructure through sales and profits.
This week’s earnings reports from U.S. tech giants are also drawing attention. Meta and Alphabet are set to release their third-quarter results on the 28th, while Amazon will follow on the 29th. The key focus will be on how much revenue these companies have generated and profits they have made from their significant investments in AI infrastructure. The results will help gauge whether AI investments are based on sustainable business demand and whether these companies have the capacity to continue large-scale investments in the future.
Amid expectations for growth in the AI industry, concerns about investment overheating are also rising. Delays in data center construction, declines in stock prices of AI infrastructure-related companies, and setbacks in some firms' IPO plans have raised questions about whether they can achieve profits commensurate with their massive investments. As the financial burden of funding infrastructure grows, there are fears that if AI business profitability does not meet expectations, the pace of investment expansion may slow.
Despite the positive earnings outlook, semiconductor stocks are showing weakness. Currently, the stock prices of Samsung Electronics and SK Hynix have fallen by more than 40% and 50%, respectively, compared to their peaks in June. Given the cyclical nature of the semiconductor industry, concerns are emerging that the current boom may be nearing its peak.
Philip Wu, portfolio manager at Relian Global Advisors, told Bloomberg that the era of unconditional optimism in the semiconductor market has ended, and scrutiny of companies' fundamentals is intensifying. He noted, "The easy money in this theme has already been made."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.