Morgan Stanley tells clients: It's time to "sell chips, buy cloud"; "storage" is like "silver" topping out.

Morgan Stanley says that AI hasn't ended, but it's now time for hyperscale cloud providers (Hyperscalers) to lead the rally. Morgan Stanley's Chief US Equity Strategist Michael Wilson issued a clear signal to clients in his latest weekly report: **Reduce semiconductor holdings, shift to hyperscale cloud providers.** This isn't bearish on AI—it's a rotation. There have been three similar adjustments during the AI investment cycle already; Wilson believes this is the fourth. After historic gains since the end of March, chip stocks have recently cooled off significantly. High-beta momentum stocks **(such as memory and chip stocks)** have recorded their largest two-day decline since the COVID-19 outbreak. Wilson judges this pullback **"may still have further room."** This judgment isn't a spur-of-the-moment call—Wilson had already proposed the "broadening trade" framework in his annual outlook in November 2025. The core logic: after rolling recessions in the US economy conclude in April 2025, a new expansion cycle will begin, earnings growth will exceed expectations, and **the leading force in the market should broaden from AI capital spending beneficiaries to a wider range of sectors.** This call was interrupted by the Iran War in February 2026. Oil prices surged, the market repriced Federal Reserve rate hikes, the broadening trade fizzled, and semiconductor stocks again were the sole outperformers thanks to the AI computing power narrative. Now, with oil prices easing and inflation expectations stabilizing, Wilson believes conditions are once again mature. --- ### Similar to “Silver Peaking”: Memory Chips are the Biggest Risk Wilson offers a concrete analogy in his report: **Semiconductors behave very much like silver.** Two reasons: first, both experienced parabolic price surges; second, both are tightly linked to commodity markets, and commodity prices are notoriously volatile. Morgan Stanley first proposed this analogy in early June, and it now appears to be coming true. He further notes that this round of adjustment will be led downward by the memory sub-sector—because memory is the most "commodity-like" part of the semiconductor complex, with high price elasticity and quick reversals. After Micron's earnings report, semiconductor stocks saw noticeable declines, which Wilson believes confirms the market's shift to focus on the "peak rate of change in earnings expectation revisions." --- ### Meta’s Statement Lit the Fuse The direct catalyst for this rotation was a Meta announcement. Last week, Meta said it will start selling its surplus computing power to external customers. This move signaled to the market that: **The pace of capital expenditure growth at hyperscale cloud providers may be reaching a turning point.** Wilson writes in his report that performance between hyperscale cloud providers (Microsoft, Google, Amazon, Meta, etc.) and semiconductor stocks has diverged sharply, but this is unsustainable—because chipmakers’ demand fundamentally depends on cloud providers’ willingness to invest. History shows that whenever this divergence becomes extreme, we see “mean reversion”: **Cloud providers either lower capex guidance or announce a shift in direction, triggering corrections in chip stocks.** Meta has just given such a reason. It’s important to emphasize that Wilson makes clear this does not mean the AI capex cycle is over, but mid-cycle resets and rotations are meaningful. His original words: “This is a peak in the rate of change of the correction, not the peak in the overall capex cycle.” In fact, since ChatGPT’s launch in November 2022, such mid-cycle corrections have now happened three times—this is time number four. --- ### Why Buy Cloud Now, Not Chips? Wilson’s reasoning: Hyperscale cloud providers (like Amazon AWS, Microsoft Azure, Google Cloud, etc.) have lagged in stock performance recently, but their fundamentals remain strong. He lists three reasons: **First, core business is robust.** Cloud providers' basic operations are strong, and they aren't entirely dependent on the AI capex narrative. **Second, unique position in AI application layer.** Wilson believes cloud providers have a leading advantage in developing and implementing the “agentic application layer,” a value that is underestimated by the market. **Third, cost-cutting leverage is overlooked.** Wilson calls this "an underestimated cost-cutting lever." Meanwhile, Morgan Stanley tracks the “high capex/sales factor,” which after strong performance over the past year, now shows signs of peaking. Cloud providers have already gone through a bout of underperformance, digesting this pressure; chip stocks may just be beginning. --- ### Broadening Market Rally: Not Just Buying Cloud Besides hyperscale cloud providers, Wilson lists other preferred directions for a broad-based rally: **Discretionary goods** are Wilson’s top pick. The logic: consumer spending is shifting from services to goods, pricing for goods is improving, and coupled with strong EPS revisions. He sees this as the "most convincing expression" of the broad earnings story. **Transports** are also beneficiaries, given the economic expansion cycle. **Biotech** represents the rate-sensitive sector. Historical data shows that in a high-and-falling rate environment, Biotech’s annualized returns approach 20%. Morgan Stanley expects core CPI to remain below 3%, forecasts for policy rates are still too hawkish, and a correction would directly benefit biotech. Continued M&A cycles also provide extra catalysts. --- ### Macro Background: Falling Oil Prices Stabilize Rates, Creating Fertile Ground for Rotation Wilson’s broadening logic also has key macro support: a sharp drop in oil prices. Falling oil prices help stabilize bond yields, and rate stability is a key driver for broadening rallies. Morgan Stanley’s base forecast: falling energy prices, tariff-driven inflation peaking, and manageable service/housing inflation together mean the Fed will keep rates unchanged this year, instead of hiking. Currently, the bond market still prices in 1.5 rate hikes before Q1 next year. Wilson says once this overly hawkish expectation is corrected, it will be a positive surprise for stocks. He also points out Fed Chair Walsh’s statement at the Sintra conference that "inflation risks have decreased," and the emphasis on the dual mandate of jobs and prices. Coupled with last week's weaker-than-expected nonfarm payroll data, Wilson believes this will further dampen hawkish rate expectations and support the broadening trade. --- ### This is Rotation, Not an End Wilson sums up at the end of the report: > “The market is starting to broaden, and the index is entering a consolidation/correction phase. This is happening now.” > > “Among AI winners, the leading sectors have rotated for years. This is just the next rotation in the cycle.” > > **“This is simply the next rotation—from semiconductors to hyperscale cloud providers, as well as the other broadening trade targets above.”** Micron’s earnings, followed by relative underperformance of semiconductors, made the market aware that "peak rate of correction" is now a core issue. Meta’s surprise announcement to sell surplus compute power validated this expectation. The consolidation of the high capex/sales factor may drive other cloud providers to lower capex guidance. All of these are fueling the broadening rally. Risk Warnings and Disclaimer The market involves risks, and investment requires caution. This article does not constitute personal investment advice, nor does it take into account individual users’ specific investment objectives, financial situations, or needs. Users should consider whether any opinions, views, or conclusions in this article fit their own situation. Investments made accordingly are their own responsibility.