Micron, SanDisk, and Memory Chips Could Face a Big Hit as Goldman Warns of Risks

Goldman Sachs warns Micron of downside risks from slowing HBM prices and Chinese competition, challenging the sustainability of the AI memory boom.

Micron and other memory-related stocks came under renewed pressure as investors reassessed the sustainability of the AI-driven memory boom, following a warning from Goldman Sachs that the sector faces several downside risks.

The weakness followed a sharp pullback in technology stocks last week, which Goldman Sachs trader Ippei Yamaura described as a rare four-standard-deviation sell-off. The initial pressure came from declines in Korean memory names before spreading to the broader AI infrastructure sector. Although Micron's latest earnings briefly supported market sentiment, the rebound failed to extend, with memory stocks selling off again after the U.S. market opened.

Goldman Sachs said the market appears to be increasingly pricing in a potential peak in memory supply tightness. The recent divergence in the technology sector also reflected that shift, with AI infrastructure stocks under pressure while hyperscale cloud providers and Apple showed relative resilience.

According to Yamaura, the correction was partly driven by renewed concerns over AI monetization and infrastructure spending. OpenAI's June 11 price cuts, aimed at competing with Anthropic, raised investor concerns that the company's path toward self-sustaining revenue could take longer than expected. Reports that OpenAI may delay its IPO until 2027 and scale back its $1 trillion valuation target further weighed on market confidence.

Goldman Sachs identified three main downside risks for Micron and the broader memory sector.

The first is slowing price momentum in high-bandwidth memory, or HBM. HBM has been a key driver of margin expansion for memory suppliers, supported by strong AI server demand and tight supply. However, Goldman warned that as industry capacity expands in fiscal years 2027 and 2028, price support could face pressure.

The second risk is rising competition from Chinese memory manufacturers. According to reports citing the Financial Times, Apple is seeking U.S. government approval to purchase DRAM chips from China's ChangXin Memory Technologies, or CXMT, in an effort to reduce cost pressure from rising memory prices. Goldman said such a move would pose a risk to Micron, particularly if major downstream customers begin seeking alternative suppliers.

The third risk is a sudden slowdown in AI server investment. Goldman noted that some customers are already looking to reduce their dependence on high-cost memory components. Qualcomm is reportedly reducing its reliance on HBM, while Nvidia is working to lower memory usage in AI systems. These moves suggest that rising memory costs may be prompting customers to seek efficiency gains or alternative system designs.

The broader concern is that elevated memory prices, while positive for supplier earnings, could pressure the economics of AI data center investment. If infrastructure costs continue to rise, hyperscalers and AI companies may face weaker returns on spending, which could eventually slow the pace of server investment.

The OpenAI developments have added to those concerns. Its reported IPO delay, lower valuation expectations, and recent price cuts have revived questions over whether AI companies can generate sufficient revenue to support the current level of infrastructure spending. That uncertainty has weighed most heavily on AI infrastructure suppliers, while companies with stronger balance sheets and more diversified revenue streams have held up better.

Despite the near-term pressure, Goldman Sachs did not frame the weakness as a structural breakdown in the broader equity market. Yamaura noted that U.S. economic fundamentals remain broadly resilient, suggesting that any rotation toward defensive assets or broader equity weakness may be cyclical rather than structural.

For Micron, the key debate is now shifting from earnings momentum to the durability of the memory cycle. The company remains a major beneficiary of AI-related demand, particularly through HBM. However, investors are increasingly focused on whether pricing strength can continue as supply expands, Chinese competition rises, and customers seek ways to lower memory costs.

Goldman's warning suggests that the memory trade may face a more difficult phase after a strong rally. If HBM tightness remains intact and AI server investment continues to grow, pullbacks in Micron and related stocks could remain attractive for long-term investors. However, if the market becomes more confident that memory pricing has peaked or that AI infrastructure spending is slowing, the sector could face further downside pressure.