--- title: "Morgan Stanley Warns: AI Memory Boom Nears Inflection Point, Memory Prices May Peak in Q4" type: "News" locale: "en" url: "https://longbridge.com/en/news/293708351.md" description: "Shawn Kim, Head of Asia and Europe Technology Research at Morgan Stanley, issued a warning: the AI-driven storage industry boom is approaching an inflection point. Memory contract prices are expected to peak in the fourth quarter, and the market's rate of earnings upgrades for storage manufacturers has slipped from a peak of 92% to 77%, with valuations of SK Hynix and Samsung Electronics both experiencing retracements. However, the cycle will \"lengthen\" rather than \"collapse\"—hard constraints on HBM supply and AI capital expenditures remain key supports, while the fortunes of commodity DRAM and high-end storage are accelerating their divergence" datetime: "2026-07-24T06:02:06.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/293708351.md) - [en](https://longbridge.com/en/news/293708351.md) - [zh-HK](https://longbridge.com/zh-HK/news/293708351.md) --- # Morgan Stanley Warns: AI Memory Boom Nears Inflection Point, Memory Prices May Peak in Q4 According to Morgan Stanley's latest report, the semiconductor storage industry boom driven by artificial intelligence (AI) is approaching an inflection point. Memory contract prices are expected to peak in the fourth quarter, marking a profound shift in the current industrial cycle. According to Zhuifeng Trading Desk, Shawn Kim, Head of Asia and Europe Technology Research at Morgan Stanley, stated in a report released on July 21 that **the momentum for earnings upgrades for storage manufacturers is significantly weakening, with data showing the net earnings upgrade rate falling from a peak of 92% to 77%.** This early signal indicates that the earnings upgrade cycle is losing momentum, and the market's pricing of storage industry earnings is returning to rationality rather than remaining in a previous state of frenzy. Affected by this, the valuations of industry giants have retraced significantly from recent highs. SK Hynix's one-year forward earnings per share (EPS) has recently declined, with its price-to-book ratio falling to 2.5x, while Samsung Electronics' price-to-book ratio has dropped to 1.7x. Although valuations remain above long-term averages, investors have begun to react to peaking growth rates and potential profit contractions. Meanwhile, inventory levels for DRAM and NAND rose in the second quarter, primarily driven by memory module manufacturers. Although the rate of cyclical change is peaking, Morgan Stanley analysts believe that this storage cycle will lengthen rather than collapse directly. ## AI Capital Expenditures Remain Strong, But Do Not Directly Imply Sustained Rise in Storage Prices Investment in AI infrastructure is the most positive of Morgan Stanley's three core judgments and the main basis for supporting the lengthening of this storage prosperity cycle. The report's focus is not just on the number of model releases, but on the training and inference intensity of leading large models, funding sources, annualized recurring revenue, and the capital expenditure trends of hyperscale cloud providers. Among these, hyperscale capital expenditures in the second quarter of 2026 are seen as a more critical verification node. However, there is an important logical gap between strong AI demand and the sustained rise in storage prices. The monetization of infrastructure does not equate to an excess of computing power; improvements in model capabilities and price declines do not automatically mean that cloud providers will indefinitely revise up their capital expenditures. **What truly determines the slope of storage demand is whether cloud providers continue to invest funds in computing power, networks, and data center infrastructure.** ## Signals of Price Inflection Point Appear First in Inventory and Earnings Expectations The judgment that storage prices will peak in the fourth quarter of 2026 is not based solely on extrapolating price curves, but on the corroboration of signals from multiple dimensions. **First is the decline in year-over-year price growth.** The year-over-year growth rate of DRAM contract prices has fallen from cyclical highs, while the market's forward price-to-book ratio for storage stocks has not expanded significantly in sync, implying that capital is not pricing the industry on the premise of a "new round of permanently high profits." **Second is the directional change in inventory.** Inventories of DRAM and NAND both increased in the second quarter, mainly driven by large module manufacturers. While rising inventory does not necessarily mean deteriorating downstream demand, against the backdrop of declining year-over-year price growth, rising inventory at the module level amplifies market sensitivity to subsequent destocking pressures. **Third is the attenuation of earnings expectation momentum.** Although the proportion of net upward earnings revisions remains in positive territory, it has dropped from a peak of 92% to 77%, narrowing the space for further upward revisions in consensus expectations. The recent decline in SK Hynix's EPS for the next year indicates that the market has begun to digest the possibility of slowing earnings growth in the next phase. When these three signals appear simultaneously, the focus of market trading often shifts from "how much more can profits rise" to "how long can profit growth be sustained"—this is a typical path where storage stocks face valuation pressure first, even while fundamentals remain healthy. ## Long-Term Agreements Have Not Led to Valuation Re-rating; Samsung and SK Hynix Still Above Long-Term Averages Long-term agreements (LTAs) signed between storage manufacturers and downstream customers are one of the key bases for Morgan Stanley's neutral stance among its three core judgments. **The market has not assigned a significantly higher valuation center to the industry due to the existence of these agreements.** This logic has historical precedents from the pandemic period: long-term agreements do not naturally eliminate cyclical risk. When prices and supply-demand relationships change, agreements may be renegotiated, or lead to passive inventory accumulation by customers. Structural demand enhancement and traditional cyclical fluctuations can coexist; the two are not mutually exclusive. In terms of valuation levels, Samsung's price-to-book ratio is approximately 1.7x, and SK Hynix's is approximately 2.5x. Both have retraced significantly from recent highs but remain above their respective long-term averages. This pattern precisely corresponds to the meaning of a neutral judgment: the industry is not viewed as purely cyclical, but the narrative of "AI reshaping storage and comprehensive valuation re-rating" has not yet been established. ## HBM Supply Tightness Remains a Hard Constraint; Commodity DRAM and High-End Storage Diverge The differentiation in supply structure is bringing two completely different supply-demand logics to the storage industry. Samsung, SK Hynix, and Micron are shifting more capacity resources to High Bandwidth Memory (HBM), objectively leaving room for commoditized DDR5; however, the tight supply situation of HBM itself has not been alleviated thereby. Commodity DRAM faces pressure from catching up with new supply, while High Bandwidth Memory remains constrained by hard constraints such as advanced processes, advanced packaging, and bandwidth capabilities. The two markets are moving towards different supply-demand equilibrium paths. High storage prices and limited bandwidth also constitute the commercial foundation for a new round of storage innovation. Morgan Stanley estimates the total addressable market for related sectors to be approximately $25 billion, covering multiple technical paths revolving around capacity, bandwidth, power consumption, and system architecture, rather than a single chip category. Long-term demand on the equipment side is closely related to the delivery pace of EUV lithography machines. Morgan Stanley predicts that ASML's EUV shipments will rise from 92 units in fiscal year 2027 to 104 units in fiscal year 2028, corresponding to expected equipment demand for the continuous expansion of advanced logic and advanced storage manufacturing, but this does not mean that all storage manufacturers will benefit synchronously. ### Related Stocks - [SKHY.US](https://longbridge.com/en/quote/SKHY.US.md) - [SSNGY.US](https://longbridge.com/en/quote/SSNGY.US.md) - [SKUU.US](https://longbridge.com/en/quote/SKUU.US.md) - [07347.HK](https://longbridge.com/en/quote/07347.HK.md) - [09347.HK](https://longbridge.com/en/quote/09347.HK.md) - [07747.HK](https://longbridge.com/en/quote/07747.HK.md) - [09747.HK](https://longbridge.com/en/quote/09747.HK.md) - [SMSN.UK](https://longbridge.com/en/quote/SMSN.UK.md) - [DRAM.US](https://longbridge.com/en/quote/DRAM.US.md) - [03121.HK](https://longbridge.com/en/quote/03121.HK.md) - [MS.US](https://longbridge.com/en/quote/MS.US.md) - [ASML.US](https://longbridge.com/en/quote/ASML.US.md) - [MS-O.US](https://longbridge.com/en/quote/MS-O.US.md) - [MS-Q.US](https://longbridge.com/en/quote/MS-Q.US.md) - [MS-E.US](https://longbridge.com/en/quote/MS-E.US.md) - [MS-I.US](https://longbridge.com/en/quote/MS-I.US.md) - [MS-L.US](https://longbridge.com/en/quote/MS-L.US.md) - [MS-P.US](https://longbridge.com/en/quote/MS-P.US.md) - [MS-A.US](https://longbridge.com/en/quote/MS-A.US.md) - [MS-F.US](https://longbridge.com/en/quote/MS-F.US.md) - [MS-K.US](https://longbridge.com/en/quote/MS-K.US.md) ## Related News & Research - [Some Investors Believe SK Hynix Could Be the Next Big AI Stock. History Says It's Not Time to Buy Shares Just Yet.](https://longbridge.com/en/news/293725955.md) - [Samsung weighs €1B Mistral stake as AI chip race tightens](https://longbridge.com/en/news/293551789.md) - [JPMorgan Sees Recent AI Selloff as a Setup for Semiconductor Sector Upswing](https://longbridge.com/en/news/293197181.md) - [Tech titan’s ex-wife fails to win greater share of AI boom in decade-long divorce saga | CNN Business](https://longbridge.com/en/news/293706827.md) - [Samsung, Broadcom sign MOU to expand memory and foundry collaboration for AI chips](https://longbridge.com/en/news/293814274.md)