Memory stocks "de-crowded"
The reason why memory stocks are among the top decliners is that they are exposed to several highly sensitive variables at the same time: AI server demand, high-bandwidth memory supply and demand, traditional dynamic random access memory and flash memory price cycles, competition among Korean storage giants, and technical selling pressure brought about by leveraged funds and ETF rebalancing. On July 16, Eastern Time, the U.S. stock market once again showed significant differentiation.
The reason why memory stocks are among the top decliners is that they are exposed to several highly sensitive variables at the same time: AI server demand, high-bandwidth memory supply and demand, traditional dynamic random access memory and flash memory price cycles, competition among Korean storage giants, and technical selling pressure brought about by leveraged funds and ETF rebalancing. On July 16, Eastern Time, the U.S. stock market once again showed significant differentiation. The Dow Jones Industrial Average fell 105.67 points, or 0.20%, to 52552.97 points; the S&P 500 fell 38.63 points, or 0.51%, to 7533.77 points; the Nasdaq Composite Index, which is dominated by technology stocks, fell 387.28 points, or 1.47%, to 25881.95 points. Compared with the index's decline, what deserves more attention is the internal structure: there are not many rising stocks among the S&P 500 constituent stocks, but the decline in semiconductor, AI hardware and memory storage stocks has a larger weight, dragging the main index lower. The memory and data storage sectors have become the focus of this round of corrections. SanDisk, Western Digital, Seagate Technology, Micron Technology and other stocks have suffered consecutive declines this week. On July 16, SanDisk fell by about 13%, Seagate Technology and Western Digital fell by more than 9%, and Micron Technology fell by about 5% to 6%. The Philadelphia Semiconductor Index fell about 4.3% that day, and BlackRock's Semiconductor ETF also fell sharply. This round of selling pressure is not limited to a single company, but a general decline. The US media generally interprets this round of decline as a "cooling of AI trading" rather than a sudden deterioration in the fundamentals of the memory industry. The Wall Street Journal stated that the decline in chip stocks reflected "the weakening of AI trading momentum." "Barron's" also pointed out that investors are beginning to question how long the AI capital spending boom can last, especially in the context of the huge gains in related stocks this year. In the past year, memory stocks have risen sharply due to expected price increases in AI servers, high-bandwidth memory, dynamic random access memory, and flash memory. Micron Technology, SanDisk, Western Digital and other companies once became the core beneficiaries of AI infrastructure investment. But when the market recently entered the earnings reporting season, the market was no longer satisfied with the narrative of "strong demand for AI" and began to ask: Can the large-scale capital expenditures of cloud vendors be converted into high enough revenue and profits? Can memory price increases continue? Will the expansion of production bring new supply pressure? The Philadelphia Semiconductor Index is an important indicator for observing global semiconductors. On July 16, the index fell about 4.3%. Data from financial data company FactSet shows that the semiconductor sector has experienced a significant correction since July. Previously, "Barron's" also mentioned that the Philadelphia Semiconductor Index has fallen by about 16% from its recent closing high. Most of the 30 constituent stocks fell, and only a few leading stocks remained relatively strong. A series of manifestations indicate that market pressure does not simply come from individual bad news of a certain memory stock, but from the overall retreat of the semiconductor sector from crowded trading. The reason why memory stocks are among the top decliners is that they are exposed to several highly sensitive variables at the same time: AI server demand, high-bandwidth memory supply and demand, traditional dynamic random access memory and flash memory price cycles, competition among Korean storage giants, and technical selling pressure brought about by leveraged funds and ETF rebalancing. From the perspective of market structure, there is currently no comprehensive risk clearing in U.S. stocks. On July 16, stocks in the medical, transportation, and some financial sectors still rose, with companies such as Abbott, UnitedHealth, and Hunter Transportation Services performing relatively strongly. This means that funds are not completely withdrawing from U.S. stocks, but are moving from the high-volatility and high-growth AI hardware chain to sectors with greater profit certainty or less valuation pressure. Currently, Wall Street institutions are clearly divided in their judgments on memory stocks, but they are not unanimously bearish. "Market Watch" pointed out that Micron Technology has become an extremely critical stock in the market because it is at the intersection of the AI supply chain and the memory price increase cycle. Although Micron Technology's stock price has fallen significantly from its highs recently, some analysts believe that its valuation based on normalized earnings is still not extremely expensive. Institutions such as Evercore still believe that tight supply of dynamic random access memory and flash memory may continue into next year, which will support prices and earnings expectations.
"Barron's" mentioned that despite Micron Technology's recent continued decline, Wall Street's average target price is still significantly higher than the current stock price, indicating that sellers have not completely given up on the upward judgment of the memory cycle. Risks also exist. Improvements in advanced equipment and process efficiency in the future may accelerate the release of supply, making future price cycles face uncertainty. Looking around the world, the decline in memory stocks has spread widely. The Associated Press reported that the decline in AI and chip stocks dragged down global markets, and South Korea's stock market was particularly under pressure. The Korea Composite Stock Price Index has been significantly affected by the decline of heavyweight stocks such as Samsung Electronics and SK Hynix. As a high-bandwidth memory core supplier, SK Hynix's fluctuations directly affect the sentiment of the global memory industry chain. In the Japanese market, semiconductor equipment and storage-related companies have also been affected. In Taiwan, China, although TSMC announced strong results, a substantial year-on-year increase in profits, and increased its capital expenditure plan, its stock price still fell. The market currently does not only look at whether financial reports are "good", but requires AI industry chain companies to prove sustainable growth and verifiable returns on capital expenditures. In addition to industry factors, the macro environment has also intensified the pressure on technology stocks. U.S. retail sales increased by 0.2% month-on-month in June, showing that consumption is still resilient; at the same time, Dallas Fed President Lori Logan issued a hawkish signal, saying that if inflation cannot sustainably return to 2%, moderate interest rate hikes may still be needed. U.S. bond yields remain high, putting highly valued growth stocks under greater pressure on discount rates. In addition, the situation in the Middle East and the risks in the Strait of Hormuz pushed up oil price fluctuations, and Brent crude oil was once at a high level. If oil prices continue to rise, it may re-raise inflation expectations and limit the Fed's room to turn to easing. This environment is not friendly to AI and semiconductor stocks that rely on future profit expectations to support their valuations. (
"Barron's" mentioned that despite Micron Technology's recent continued decline, Wall Street's average target price is still significantly higher than the current stock price, indicating that sellers have not completely given up on the upward judgment of the memory cycle. Risks also exist. Improvements in advanced equipment and process efficiency in the future may accelerate the release of supply, making future price cycles face uncertainty. Looking around the world, the decline in memory stocks has spread widely. The Associated Press reported that the decline in AI and chip stocks dragged down global markets, and South Korea's stock market was particularly under pressure. The Korea Composite Stock Price Index has been significantly affected by the decline of heavyweight stocks such as Samsung Electronics and SK Hynix. As a high-bandwidth memory core supplier, SK Hynix's fluctuations directly affect the sentiment of the global memory industry chain. In the Japanese market, semiconductor equipment and storage-related companies have also been affected. In Taiwan, China, although TSMC announced strong results, a substantial year-on-year increase in profits, and increased its capital expenditure plan, its stock price still fell. The market currently does not only look at whether financial reports are "good", but requires AI industry chain companies to prove sustainable growth and verifiable returns on capital expenditures. In addition to industry factors, the macro environment has also intensified the pressure on technology stocks. U.S. retail sales increased by 0.2% month-on-month in June, showing that consumption is still resilient; at the same time, Dallas Fed President Lori Logan issued a hawkish signal, saying that if inflation cannot sustainably return to 2%, moderate interest rate hikes may still be needed. U.S. bond yields remain high, putting highly valued growth stocks under greater pressure on discount rates. In addition, the situation in the Middle East and the risks in the Strait of Hormuz pushed up oil price fluctuations, and Brent crude oil was once at a high level. If oil prices continue to rise, it may re-raise inflation expectations and limit the Fed's room to turn to easing. This environment is not friendly to AI and semiconductor stocks that rely on future profit expectations to support their valuations. (