Morgan Stanley Research Report Interpretation: Storage Cycle Enters Second Half, Market Is Switching Pricing Anchors
TechFlow Selected TechFlow Selected
Morgan Stanley Research Report Interpretation: Storage Cycle Enters Second Half, Market Is Switching Pricing Anchors
Storage is still in the mid-to-late stage of the upward cycle, but the logic driving stock prices has already shifted from "price increase elasticity" to "earnings sustainability".
Navigating Web3 tides with focused insightsBy: Rita
TechFlow Guide
Over the past month, Asian memory stocks fell 15% to 25%, while the chip sector as a whole traded sideways. Memory stocks were sold off unilaterally, and the valuation premium between the two quickly compressed from over 1.5x to 1.1x.
Morgan Stanley gave its judgment in the TMT webcast on July 14: memory fundamentals have not reversed; the market is completing a switch in pricing anchor.
In the first half of the year, everyone cared about "how high prices can rise," while in the second half, the question shifted to "how long high profits can last." Three points of contention are influencing this judgment: how much CSP capital expenditure will increase, whether LTA is protecting profits or limiting profits, and whether YMTC will break the supply-demand balance of NAND.
Morgan Stanley's conclusion is: memory is still in the mid-to-late stage of the upcycle, but the logic driving stock prices has switched from "price increase elasticity" to "profit sustainability."
What Happened in the Past Month
Chip stocks fell 15% to 25% while the broader market traded sideways, the SOX index basically didn't move, but memory companies were sold off unilaterally.
Fundamentals have not reversed; it is the valuation framework that is moving. The valuation premium of memory over SOX quickly compressed from the high point to 1.1x, returning to the upper edge of the historical reasonable range. The market is switching from "chasing price increases" to "verifying profit sustainability," which is a typical characteristic of the cycle entering the mid-to-late stage.
Demand for memory from AI servers is still increasing, HBM capacity remains tight, and the supply-demand gap for traditional DRAM and NAND has not been fully closed. However, the market is no longer willing to pay unlimited premiums for "price increase expectations"; it is starting to demand evidence that "profits remain stable after price increases."
Controversy One: How Much Will CSPs Spend
AI servers are the ultimate source of memory demand. The CSP capital expenditure number is the most important question.
Morgan Stanley estimates total CSP hardware capital expenditure will be about $339 billion in 2026 and about $406 billion in 2027, 30% and 37% higher than market consensus respectively. This is the biggest point of divergence between the market and sell-side. Why the large difference? Land reserves, power approvals, and number of projects under construction for hyperscale cloud vendors are all at historical highs; these leading indicators point to capital expenditure not peaking in 2026.
Whether this expectation can be realized is the biggest variable in the short term. The earnings season starting at the end of July will provide the first verification. If CSPs raise capital expenditure, market confidence in long-term memory demand will strengthen. If the numbers are below expectations, the market will question the sustainability of the cycle.
Controversy Two: Is LTA Protecting Profits or Limiting Profits
LTA is the topic with the greatest divergence among institutional investors.
Concerns focus on: after a large number of contracts are locked, price elasticity will decline, and short-term EPS elasticity brought by price increases will be compressed. This concern is reasonable from a short-term perspective, but the goal of this cycle has shifted from rising higher to lasting longer.
Morgan Stanley's view is the opposite. LTA provides a floor for profits, not a cap for prices. In the no-agreement era, memory manufacturers faced the worst-case scenario of plummeting prices during cycle downturns. After the introduction of LTA, more than half of contract volume is locked within agreement price ranges, and the risk of falling through during downturns is significantly reduced.
The market's past valuation habit for memory stocks was built on the assumption that "prices will fluctuate violently." LTA is loosening this assumption. Once LTA covers most contract volume, the fear of cycle downturns will be significantly weakened. Morgan Stanley believes the market has not fully priced in this change, but it is happening.
Controversy Three: Will YMTC Break the Balance of NAND
YMTC is the representative of China's NAND capacity expansion. Morgan Stanley conducted a scenario analysis: Fab4 and Fab5 each plan about 100kwpm capacity; after all five fabs are put into production, theoretical capacity could reach 24% of global market share.
The key lies in the method and timing of capacity release, not the numbers themselves. If YMTC puts into production gradually according to market demand and does not expand aggressively, NAND supply-demand tightness can last until 2028. If all five fabs are put into production as soon as possible, the NAND market will face severe oversupply, and the price system may be broken.
Morgan Stanley believes there is currently no evidence that YMTC is accelerating capacity expansion. But this is a variable that needs continuous tracking. If Chinese enterprises obtain more advanced technology nodes, or the external environment changes, the speed of capacity release may change the supply-demand pattern of the entire market.
TechFlow Perspective
The pricing logic of memory stocks is switching from "cyclical products" to "stable profit products." The habit of the past ten years was "give high prices when rising, give floor prices when falling," because everyone knew rises and falls would be violent. But the popularity of LTA is flattening the amplitude. Once more than half of contracts are locked by agreement, the fear of cycle downturns will disappear.
What Morgan Stanley was actually saying in the webcast on July 14 is one thing: the memory cycle is still alive, but the driving force has changed. The first half is "price increase elasticity," the second half is "profit sustainability." CSP determines the demand ceiling, LTA determines the profit floor, and YMTC determines supply risk.
Three variables point in one direction simultaneously: the downside risk of memory stocks is smaller than historical cycles, and the upside elasticity is also weaker than historical cycles. The market's divergence lies in whether this combination is bullish or bearish. Morgan Stanley's view is clear: memory stocks are switching from "cyclical high volatility" to "structural medium-to-high returns," and the market has not fully realized that this change means a fundamental change in valuation methods.
The question now is, how many companies can pass this test.
Disclaimer
This article is organized and interpreted by TechFlow Research based on a third-party brokerage research report (Morgan Stanley, July 14, 2026). The ratings, target prices, earnings forecasts, and related judgments cited in the text are the views of the brokerage analysts, represent only their affiliated institution's position, do not represent TechFlow Research's views, and do not constitute any investment advice.
The market has risks, decisions need independence. This article should not be used as a basis for buying or selling any securities.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@chaoxiangooo
