--- title: "Trillion-Dollar Frenzy for Memory Sellers, Halved Profits for Memory Buyers" type: "News" locale: "en" url: "https://longbridge.com/en/news/287756039.md" description: "Xiaomi released its Q1 2026 financial report, with total revenue of RMB 99.1 billion, a year-on-year decrease of 10.9%, and net profit plunging 43.1%. Memory prices surged nearly fourfold, leading Xiaomi to cut entry-level models and reducing quarterly shipments to 33.8 million units. Meanwhile, Micron Technology's market capitalization exceeded $1 trillion, with its stock price rising eightfold in one year. Goldman Sachs remained cautious during this rally, maintaining a neutral rating on Micron" datetime: "2026-05-27T10:38:38.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/287756039.md) - [en](https://longbridge.com/en/news/287756039.md) - [zh-HK](https://longbridge.com/zh-HK/news/287756039.md) --- # Trillion-Dollar Frenzy for Memory Sellers, Halved Profits for Memory Buyers Two significant events occurred simultaneously on the evening of May 26. Xiaomi released its financial report for the first quarter of 2026. Total revenue was RMB 99.1 billion, down 10.9% year-on-year; adjusted net profit was RMB 6.07 billion, plummeting 43.1% year-on-year. Revenue from the smartphone business was RMB 44.3 billion, down 12.5% year-on-year, with gross margin falling to 10.1%, a decrease of 2.3 percentage points compared to the same period last year. During the earnings call, Lu Weibing, President of XIAOMI-W, cited a figure: memory prices for the same specifications surged nearly fourfold compared to the same period last year. For a smartphone with 12GB LPDDR5 + 512GB UFS configuration, the memory cost alone increased by approximately RMB 1,500. He stated that Xiaomi "would not pass on the cost of memory price hikes to consumers," but also predicted that the price increase cycle would continue into 2027 or even 2028. To survive, Xiaomi proactively cut entry-level models, causing quarterly shipments to drop to 33.8 million units. The second event was that Micron Technology surged more than 19% in a single day, with its market capitalization breaking through $1 trillion. UBS raised Micron's target price directly from $535 to $1,625, an increase of about 204% at once, making it the highest target price among the 46 brokerages currently covering Micron. Just a few days earlier, Citigroup had raised Micron's target price from $425 to $840, and HSBC also raised it from $750 to $1,100. Wall Street has rarely seen such unified opinion on a single cyclical stock. Twelve months ago, Micron's stock price was below $110. It rose eightfold within a year. **On the same day, memory sellers enjoyed a trillion-dollar frenzy, while memory buyers saw their profits halved.** Goldman Sachs played an intriguing role in this frenzy. In December 2025, Goldman Sachs gave Micron a neutral rating with a target price of $205. In the first quarter of 2026, Goldman Sachs reduced its position in Micron by nearly 20%. On March 19, the day Micron released its earnings, Goldman Sachs adjusted its target price from $360 to $400 but maintained a neutral rating, even though the stock price was already far above $400. Micron then surged 40% in a week, leaving Goldman Sachs precisely missing the rally. On May 17, Goldman Sachs issued a report on the storage industry, concluding that it was facing "the most severe supply shortage in 15 years" and upgraded the overall rating for the storage industry. However, it remained neutral on Micron, with the target price still at $400. **Goldman Sachs, as an outlier, is either the last sober person in this frenzy or the one who missed the rally most severely.** But this strong divergence is worth serious consideration. ## Why the Crazy Rise? A New Story Called LTA? In a research report on May 26, UBS analyst Timothy Arcuri's core argument was that Long-Term Agreements (LTA) are fundamentally eliminating the cyclicality of the storage industry. Storage chips are the most commodity-like products in the semiconductor industry. For forty years, DRAM and NAND prices have followed a cruel pattern: two years up, two years down, with price collapses never absent. The profits of Micron, Samsung, and SK Hynix fluctuate like an electrocardiogram, and the market has never dared to value these companies based on "steady-state earnings." For forty years, the valuation fluctuation range for cyclical stocks has been roughly 8 to 15 times P/E. Chart: Micron's financial data showing electrocardiogram-like fluctuations **UBS's story is that the "cycle curse" for these companies will be broken, with "AI" as the main protagonist.** Cloud providers such as Microsoft, Google, Amazon, and Meta, in order to secure HBM and DDR5 supply in the AI arms race, have begun to actively sign 3-to-5-year fixed-price long-term contracts with storage manufacturers, including prepayments. These contracts are not the "indicative" agreements common in the traditional semiconductor industry, but binding procurement commitments that lock in volume, price, and even wafer capacity. Chart: AI Capital Expenditure of Major Tech Companies (2022—2026E): The combined total for the four companies is expected to reach $725 billion in 2026. Individually, Amazon $200 billion, Microsoft $190 billion, Alphabet $190 billion, and Meta $145 billion. The 2026 data represents the latest guidance upper limits as of April 29 for each company, with Microsoft's figures based on calendar year totals derived from quarterly data. Microsoft and Google were reported in April to be negotiating three-year long-term DRAM contracts with SK Hynix, structures that include upfront deposits. Previously, manufacturers begged customers to place orders; now, customers pay deposits to lock in capacity. The power dynamics in the supply chain have reversed. UBS's model calculations show that if LTAs are incorporated into Micron's earnings forecasts, Micron's annual earnings per share (EPS) could remain above $100 even if DRAM spot prices plummet by 50% in fiscal year 2029. LTAs can narrow the volatility amplitude of DDR prices from peak to trough by about 50%. By 2027, 20% to 30% of the industry's total DDR bit shipments will be locked in by fixed-price long-term agreements. Among the DDR5 purchases by top hyperscalers, 60% to 70% may already be under fixed contracts. From a valuation perspective, if cyclicality disappears, storage stocks should not be valued as cyclical stocks but as infrastructure utilities, jumping from 8-15x P/E to 20-30x P/E. JPMorgan also issued a research report with similar conclusions in mid-May, titled "LTA is Eliminating Cyclicality in the Storage Industry." Citigroup's logic is that HBM production will squeeze out general DRAM wafer capacity, leading to long-term shortages in general storage as well. Micron's stock price surge has welcomed a Davis double play of switching profit and valuation systems. ## This Storage Is Not That Storage Wall Street uses the term "storage supercycle" to tell a unified bull market narrative. But "storage" and "storage" are completely different. The storage market in 2026 presents a three-tier differentiation. The first tier is AI storage: HBM, server DDR5, and enterprise SSDs. Here, price hikes, shortages, and long-term contracts locking capacity are happening simultaneously. TrendForce predicts that DRAM contract prices will rise 58% to 63% quarter-on-quarter in Q2 2026, and NAND Flash contract prices will rise 70% to 75%; Kioxia also publicly stated that its 2026 capacity is basically sold out. This tier is the story behind Micron's trillion-dollar market cap. The second tier is mobile and embedded storage: mobile DRAM and mobile NAND. Prices here are also rising sharply. Counterpoint data shows that DRAM prices rose more than 50% quarter-on-quarter in Q1 2026, and NAND Flash prices rose more than 90% quarter-on-quarter. Related reports from TrendForce indicate that memory usually accounted for about 10% to 15% of a phone's BOM, but has now risen to 30% to 40%, with pressure more evident in low-end models. Left chart DRAM (memory) trend: Low-end phones saw the sharpest increase, climbing from initial low levels to a predicted 35% in Q2 2026; high-end phones to 23%; mid-range phones to 20%. The dashed part (after Q1 2026) represents forecast values. Right chart NAND (flash) trend: All price ranges were basically stable in the first three quarters of 2025, but surged sharply starting from Q4 2025. **Xiaomi is in this tier. Its pain is that "AI has grabbed capacity, leaving less for phones, and phone manufacturers must pay higher prices for the remaining capacity."** Original equipment manufacturers (OEMs) gave priority capacity to AI customers, leaving phone manufacturers with few choices in contract procurement. If you want to ship, you must buy at the new contract price; if you don't, your production lines and new product rhythms will be affected. The third tier is PC retail spot market: DDR5 modules and consumer-grade SSDs. Here, opposite fluctuations occurred. TrendForce reported that by the end of March, 32GB DDR5 modules in Chinese channels fell from nearly RMB 3,000 by RMB 500 to RMB 1,050, with some clearance prices as low as RMB 1,950; Tom’s Hardware also wrote that some DDR5 products in Chinese and overseas retail markets fell 25% to 30% from their highs. But this is mainly a split between retail spot markets and contract procurement. PC channels have inventory and can dump stock; phone manufacturers procure by contract and have no option to dump. The same "storage" industry, three tiers, three directions. The essence of this differentiation is that the three giant storage manufacturers are shifting wafer capacity from consumer-grade to AI. HBM production squeezes out general DRAM wafers, enterprise SSDs squeeze out consumer NAND supply, leaving less capacity for phones and PCs. Phone manufacturers are forced to accept price hikes because they must ship; PC channels can press prices and dump stock because inventory is sufficient. Image generated with AI assistance **Micron and others actively chose to give capacity to AI customers willing to pay more. In the short term, this is a beautiful upgrade in product structure. But it also means Micron is blocking its retreat; once AI demand slows, capacity may not switch back smoothly.** Micron's financial report shows that quarter-on-quarter, DRAM bit shipments grew only in the mid-single digits, and NAND bit shipments grew only in the low single digits, with growth mainly coming from ASP increases. Micron's story today is solely about "extreme shortage in the AI storage branch." Micron has bet everything on this branch. ## **Can Long-Term Agreements Really Eliminate Cycles?** The logic of long-term agreements seems solid. Under the spending rhythm of AI, the supply elasticity of storage chips is extremely low. It takes 18 to 24 months for HBM capacity to go from planning to production, and producing HBM squeezes out general DRAM wafers. Cloud providers sign long-term agreements because they worry about "AI project delays." **But there is a premise for long-term agreements to eliminate cyclicality: demand does not collapse.** Different institutions have different statistical scopes for AI CapEx, but the direction is consistent: AI infrastructure investment is rushing from the hundreds of billions of dollars level to nearly the trillion-dollar level. According to some market models, this is a capital expenditure curve with an annualized growth rate close to 40% to 50%. However, nothing in the physical world grows at more than 40% forever. It does not require an AI bubble to burst; just a slowdown in growth from 45% to 20% could reverse the supply-demand balance of storage chips within 18 months. The three storage manufacturers are now expanding production crazily. Micron's CapEx for fiscal year 2026 is $25 billion, with another $10 billion added in 2027. There is another matter that must be faced squarely: when a company's revenue growth relies entirely on price elasticity rather than volume elasticity, the story is fragile. Micron's shipments only increased by 4% to 6%, and the 196% revenue growth was mainly due to price hikes. Prices can go up and down, and they fall much faster than they rise. This is also the essence of cyclicality. Let's do a simple arithmetic problem. Micron's current market capitalization is $1 trillion. Micron has increased its fiscal 2026 CapEx to over $25 billion and expects capital expenditures to continue to increase significantly in fiscal 2027, with some market reports mentioning incremental amounts possibly exceeding $10 billion. Micron's non-GAAP net profit for the second quarter of fiscal 2026 was about $14 billion, simply annualized to about $56 billion, corresponding to about 18 times P/E. If subsequent price hikes and long-term agreements are further extrapolated, the P/E can still be calculated at around 15 times. It might seem "cheap." But the denominator of this P/E is the peak earnings of a supercycle where DDR4 contract prices rose 10-fold in 15 months, HBM was sold out for the whole year, and gross margin jumped from 36% to 75%. Using peak-cycle earnings multiplied by a seemingly "reasonable" multiple to derive a seemingly "not expensive" valuation is precisely the classic valuation trap when cyclical stocks peak. Cisco's P/E in 2000 was also "only" 60-something times, built on 15 consecutive quarters of 50%+ revenue growth. When growth slowed from 50% to 20% and then to 0%, EPS did not need to fall much for the stock price to drop 80%, because both the multiple and earnings contracted. From a Davis double play to a double kill. History tells us one thing: in commodity markets, long-term agreements are never a one-sided "floor." They protect buyers in upcycles and sellers in downcycles, but only if both sides have the ability and willingness to fulfill contracts. The moment long-term agreements are truly needed is precisely when they are most likely to fail. This is not to say that Micron is definitely a bubble. AI's demand for computing power and storage may indeed be structural, LTA may indeed have rewritten industry rules, and a trillion-dollar market cap may just be the starting point. But when the entire Wall Street simultaneously shouts "this time is different," it is at least worth stopping to ask: What happened the last time everyone was so certain? In a sense, one can only make money by enjoying the frenzy of the bubble. However, Cisco took about 25 years, until today's AI era, to surpass the closing high of the internet bubble period, and the internet did indeed change everything. Risk Warning and Disclaimer The market carries risks, and investment requires caution. This article does not constitute personal investment advice, nor does it take into account the specific investment goals, financial status, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investment based on this content is at the user's own risk. ### Related Stocks - [01810.HK](https://longbridge.com/en/quote/01810.HK.md) - [MU.US](https://longbridge.com/en/quote/MU.US.md) - [SSNGY.US](https://longbridge.com/en/quote/SSNGY.US.md) - [MSFT.US](https://longbridge.com/en/quote/MSFT.US.md) - [GOOGL.US](https://longbridge.com/en/quote/GOOGL.US.md) - [GOOG.US](https://longbridge.com/en/quote/GOOG.US.md) - [AMZN.US](https://longbridge.com/en/quote/AMZN.US.md) - [META.US](https://longbridge.com/en/quote/META.US.md) - [CSCO.US](https://longbridge.com/en/quote/CSCO.US.md) - [XIACY.US](https://longbridge.com/en/quote/XIACY.US.md) - [81810.HK](https://longbridge.com/en/quote/81810.HK.md) - [HXXD.SG](https://longbridge.com/en/quote/HXXD.SG.md) - [SMSN.UK](https://longbridge.com/en/quote/SMSN.UK.md) ## Related News & Research - [J.P. Morgan Keeps Their Hold Rating on Xiaomi (XIACF)](https://longbridge.com/en/news/286181271.md) - [Xiaomi EV reportedly taps new battery suppliers for second brand to target EREV market](https://longbridge.com/en/news/286843996.md) - [Finance titan Jan van Eck warns memory stocks look 'bubblelicious'](https://longbridge.com/en/news/287778448.md) - [Kuaishou beats estimates as Kling AI video tool’s revenue jumps 300%](https://longbridge.com/en/news/287745352.md) - [TRUMP SAYS GETTING ALONG REALLY WELL WITH VENEZUELA TRUMP ON HORMUZ CONTROL: THAT'S PART OF THE NEGOTIATION](https://longbridge.com/en/news/287803803.md)