Memory chip stocks, including giants like Samsung Electronics, SK Hynix, and Micron Technology, have experienced a sharp downturn, raising concerns among analysts about the sustainability of the AI memory 'supercycle.' On August 18th, Samsung Electronics and SK Hynix saw declines of 7.82% and 9.75% respectively. Micron Technology plunged 7.02% to $940.76, while its shares had previously climbed about 18% over five trading sessions. Other memory and data storage stocks like SanDisk, Western Digital, and Seagate Technology also fell, with SanDisk and Seagate dropping around 9% and Western Digital down 7% en.sedaily.com.
The recent sell-off is largely attributed to profit-taking after significant year-to-date gains, with SanDisk up 653% and Micron up 255% before the dip. A Wall Street Journal analysis revealed that nine top tech companies have approximately $3 trillion in off-balance-sheet AI commitments, a figure triple their outstanding leases and long-term borrowings. This raised market doubts about the long-term sustainability of AI investment. Despite strong fundamentals, including Micron's guidance for fiscal Q4 revenue of $50 billion ± $1 billion and an 86% gross margin, and CEO Sanjay Mehrotra stating that "DRAM and NAND industry demand continues to significantly exceed industry supply," investors are rotating out of AI hardware 247wallst.com.
Even with robust financial results, memory stocks have struggled to maintain momentum. SK Hynix, for example, reported a 257% year-over-year revenue climb and a 557% surge in operating profit, yet its stock continued to fall, dropping another 8.6% at market open. Micron also saw a 32% decline in its stock price since June, despite delivering record quarterly revenue and raising guidance above expectations. Analysts point out that the market is now less concerned with current profit highs and more focused on how long elevated profits can last 247wallst.com, 404kresearch.substack.com.
The market's shift in focus is evident in how memory companies' stock prices have reacted post-earnings. While some, like Micron and Kioxia, saw initial gains after reporting strong results, they later experienced significant declines. For instance, Micron reached a new high after earnings but then traded as much as 29.5% below its baseline. This indicates that the market is applying a two-round pricing strategy: an initial reaction to earnings beats, followed by a more sustained evaluation of profit duration, capital expenditure, and valuation. This second round of pricing has seemingly overwhelmed the first, leading to continued stock declines despite strong fundamentals 404kresearch.substack.com.
Adding to the market's unease are concerns that rising memory prices could actually squeeze the profitability of AI investments, potentially cracking the supercycle that AI initially created. As data centers demand high volumes of High Bandwidth Memory (HBM), legacy DRAM, and NAND, memory prices have surged. General DRAM prices, for example, rose 13% to 18% from the previous quarter. If memory demand slows due to factors like spiking interest rates or AI models that fall short of profitability expectations, a decline in memory prices and a correction in memory makers' earnings and share prices could be inevitable en.sedaily.com. Some experts, however, remain optimistic, asserting that HBM demand will only rise as AI applications expand across industries en.sedaily.com.
