<div id="readability-page-1"><div><div><img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7472946_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="339" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7472946_image3.png"/>Author: Xiaojing, Tencent TechnologyTwo things happened simultaneously on the evening of May 26th.Xiaomi released its Q1 2026 financial report. Total revenue was 99.1 billion yuan, a year-on-year decrease of 10.9%; adjusted net profit was 6.07 billion yuan, a year-on-year plunge of 43.1%. Xiaomi's mobile phone business revenue was 44.3 billion yuan, a year-on-year decrease of 12.5%, and the gross profit margin fell to 10.1%, a decrease of 2.3 percentage points compared to the same period last year. During the earnings call, Xiaomi Group President Lu Weibing mentioned a figure: the price of memory for the same version has surged nearly four times compared to the same period last year. For a phone with a 12GB LPDDR5 + 512GB UFS configuration, the memory cost alone is about 1,500 yuan higher. He said Xiaomi &#34;will not pass on the increased memory costs to consumers,&#34; but also predicted that the price increase cycle will continue into 2027 or even 2028. To survive, Xiaomi proactively cut entry-level models, and quarterly shipments fell to 33.8 million units. Secondly, Micron Technology's stock price surged by more than 19% in a single day, pushing its market capitalization above $1 trillion. UBS raised its target price for Micron from $535 to $1625, a one-time increase of approximately 204%, making it the highest target price among the 46 brokerages currently covering Micron. Just days earlier, Citigroup had raised its target price for Micron from $425 to $840, and HSBC also raised it from $750 to $1100. Wall Street hasn't seen such unanimous agreement on the same cyclical stock in a long time. Twelve months ago, Micron's stock price was less than $110. It has increased eightfold in a year. On the same day, the trillion-dollar frenzy of selling memory chips saw memory chip 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 Micron holdings by nearly 20%. On March 19th, the day Micron released its earnings report, Goldman Sachs raised its target price from $360 to $400, but maintained a neutral rating, even though the stock price was already well above $400. Micron then surged 40% in a week, and Goldman Sachs missed the boat. On May 17th, Goldman Sachs released a report on the storage industry, concluding that it was experiencing the &#34;most severe supply shortage in 15 years,&#34; and upgraded its overall rating for the storage industry. However, it remained neutral on Micron, with the target price still at $400. Goldman Sachs, this outlier, was either the last to wake up from this frenzy or the one who missed out the most. However, this strong divergence is worth serious consideration. Why the crazy surge? A new story called LTA? A research report by UBS analyst Timothy Alcuri on May 26th argues that Long-Term Agreements (LTAs) are fundamentally eliminating the cyclicality of the memory industry. Memory chips are the most commodity-like product in the semiconductor industry. For forty years, DRAM and NAND prices have followed a brutal pattern: two years of increases, two years of decreases, and price collapses have never been absent. The profits of Micron, Samsung, and SK Hynix resemble an electrocardiogram; the market has never dared to value these companies based on &#34;steady-state profitability.&#34; For the past forty years, the valuation fluctuation range of cyclical stocks has roughly been between 8 and 15 times price-to-earnings ratio. [Image: Micron's financial data in an electrocardiogram-like pattern] UBS's story is that this &#34;cyclical curse&#34; of these companies will be broken, and the protagonist behind it is &#34;AI&#34;. In order to secure HBM and DDR5 supplies in the AI ​​arms race, cloud providers such as Microsoft, Google, Amazon, and Meta have begun to proactively sign long-term, fixed-price contracts of 3 to 5 years with memory manufacturers, often including upfront payments. These contracts are not the &#34;letters of intent&#34; of the traditional semiconductor industry; they are binding procurement commitments that lock in quantities, prices, and even wafer capacity. <img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7472948_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="555" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7472948_image3.png"/>Chart: AI Capital Expenditure of Major Tech Companies (2022-2026E): The four companies' combined expenditure is projected to reach $725 billion in 2026. Individually, Amazon is projected at $200 billion, Microsoft at $190 billion, Alphabet at $190 billion, and Meta at $145 billion. The 2026 figures represent the latest upper limit of guidance for each company as of April 29th. Microsoft's figures are calendar year totals based on quarterly data.<img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7472948_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="555" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7472948_image3.png"/> In April, it was reported that Microsoft and Google were in talks with SK Hynix about a three-year DRAM contract, which included a prepayment deposit. Previously, manufacturers sought orders from customers; now, customers are paying deposits to lock in production capacity. The power dynamics in the industry chain have reversed. UBS's model calculations show that if LTAs are included in Micron's profit forecasts, even if DRAM spot prices plummet by 50% in fiscal year 2029, Micron's full-year earnings per share will still remain above $100. LTAs can narrow the fluctuation range of DDR prices from cycle peak to trough by about 50%. By 2027, 20% to 30% of the total DDR bit shipments in the industry will be locked in with fixed-price long-term contracts. For leading hyperscalers, 60% to 70% of their DDR5 procurement may already be under fixed contracts. From a valuation perspective, if the cyclicality disappears, storage stocks should not be valued as cyclical stocks, but rather as infrastructure utilities, jumping from a PE ratio of 8-15 to 20-30. JPMorgan Chase also released a research report in mid-May with a similar conclusion, titled &#34;LTA is Eliminating Cyclicality in the Storage Industry.&#34; Citigroup's logic is that HBM production will crowd out ordinary DRAM wafer capacity, leading to a long-term shortage of general-purpose storage. Micron's stock price surge benefited from a double boost from both profit and valuation system shifts. This storage is not the same as that storage. Wall Street uses the term &#34;storage supercycle&#34; to tell a unified bull market narrative. But not all &#34;storage&#34; is the same. The storage market in 2026 will exhibit a three-tiered differentiation. The first tier is AI storage: HBM, server DDR5, and enterprise-grade SSDs. Here, price increases, stockouts, and long-term contracts locking in production capacity occur simultaneously. TrendForce predicts that DRAM contract prices will rise by 58% to 63% quarter-on-quarter in Q2 2026, and NAND Flash contract prices will rise by 70% to 75%; Kioxia has also publicly stated that its 2026 production capacity is basically sold out. This tier is the story of Micron's trillion-dollar market capitalization. The second tier is mobile and embedded storage: mobile DRAM and mobile NAND. Price increases are also severe here. Counterpoint data shows that DRAM prices rose by more than 50% quarter-on-quarter in Q1 2026, and NAND Flash prices rose by more than 90% quarter-on-quarter. A TrendForce report shows that memory, which used to account for about 10% to 15% of a phone's BOM (Bill of Materials), has now risen to 30% to 40%, with lower-end models facing even greater pressure. The left chart shows DRAM (memory) trends: lower-end phones saw the most significant increase, climbing from an initial low level to a projected 35% by Q2 2026; high-end phones to 23%; and mid-range phones to 20%. The dotted line (Q1 2026 and beyond) represents projected values. The right-hand chart shows NAND (flash memory) trends: all prices remained relatively stable in the first three quarters of 2025, but surged from Q4 2025 onwards. Xiaomi is in this tier. Its pain point is that &#34;AI has taken away production capacity, leaving less for mobile phones, forcing mobile phone manufacturers to pay higher prices for the remaining capacity.&#34; Original equipment manufacturers (OEMs) prioritized production capacity for AI clients, leaving mobile phone manufacturers with few options for contract procurement. To ship, they had to buy at the new contract price; otherwise, production lines and new product launch schedules would be affected. The third tier is PC retail spot goods: DDR5 modules and consumer-grade SSDs. Here, the opposite trend occurs. TrendForce reports that by the end of March, the price of 32GB DDR5 modules in Chinese channels had dropped by 500 to 1050 yuan from nearly 3000 yuan, with some clearance prices as low as 1950 yuan. Tom’s Hardware also noted that some DDR5 products in the Chinese and overseas retail markets had fallen by 25% to 30% from their peak. This is mainly due to the divergence between retail spot goods and contract purchases. PC channels have inventory and can sell off their stock; mobile phone manufacturers purchase under contracts and have no option to sell off their stock. Within the same &#34;storage&#34; industry, there are three layers and three directions. The essence of this divergence is that the three major storage manufacturers are shifting wafer capacity from consumer-grade to AI. HBM production is crowding out ordinary DRAM wafers, and enterprise-grade SSDs are crowding out consumer-grade NAND supply, leaving less capacity for mobile phones and PCs. Mobile phone manufacturers are forced to accept price increases because they must ship their products; PC channels, with ample inventory, can lower prices to sell off their stock. Micron and its competitors have proactively chosen to allocate production capacity to AI customers who are more willing to pay. In the short term, this is a smart product structure upgrade. However, it also means that Micron is closing off its escape route; if AI demand slows down, production capacity may not be able to smoothly shift back. Micron's financial report shows that, sequentially, DRAM bit shipments only increased by mid-single digits, and NAND bit shipments only increased by low single digits, with the growth mainly driven by rising ASPs. Micron's story today revolves around the &#34;extreme shortage of AI storage.&#34; Micron has staked everything on this branch. Can long-term contracts truly eliminate cycles? The logic of long-term contracts seems solid. Given the pace of AI spending, the supply elasticity of memory chips is extremely low. HBM capacity takes 18 to 24 months from planning to production, and HBM production will crowd out general-purpose DRAM wafers. Cloud vendors sign long-term contracts because they are worried about &#34;AI project delays.&#34; However, long-term contracts eliminate cyclicality with one prerequisite: demand doesn't collapse. Different institutions use different statistical methods for AI CapEx, but the direction is consistent: AI infrastructure investment is surging from hundreds of billions of dollars to nearly a trillion dollars. According to some market models, this represents an annualized capital expenditure curve of nearly 40% to 50%. However, nothing in the physical world grows at over 40% forever. The AI ​​bubble doesn't need to burst; a drop in growth rate from 45% to 20% could reverse the supply-demand balance of memory chips within 18 months. All three major memory manufacturers are currently expanding production at an alarming rate, with Micron's CapEx at $25 billion in fiscal year 2026 and an additional $10 billion in 2027. There's also something we must acknowledge: when a company's revenue growth relies entirely on price elasticity rather than sales volume elasticity, the story is fragile. Micron's shipments only increased by 4% to 6%, while revenue growth of 196% was mainly driven by price increases. Prices can rise or fall, and they fall much faster than they rise. This is the nature of cyclicality. Let's do a simple arithmetic. Micron's current market capitalization is $1 trillion. Micron has increased its fiscal year 2026 CapEx to over $25 billion and expects capital expenditures to continue to increase significantly in fiscal year 2027, with some market reports suggesting an increase of over $10 billion. Micron's non-GAAP net profit for the second quarter of fiscal year 2026 is approximately $14 billion, which translates to approximately $56 billion annualized, corresponding to a PE ratio of approximately 18. If we extrapolate subsequent price increases and long-term contracts, the PE ratio can be calculated to be around 15. It might seem like a bargain. But the denominator in this PE ratio is the peak earnings of a supercycle: DDR4 contract prices increased tenfold in 15 months, HBM sold out for the entire year, and gross margin jumped from 36% to 75%. Multiplying the peak earnings by a seemingly &#34;reasonable&#34; multiple yields a seemingly &#34;inexpensive&#34; valuation, which is the classic valuation trap when cyclical stocks peak. In 2000, Cisco's PE ratio was &#34;only&#34; around 60, based on 15 consecutive quarters of revenue growth of 50%+. When the growth rate dropped from 50% to 20% and then to 0%, EPS didn't need to fall much, but the stock price could drop by 80% because both the multiple and earnings contracted simultaneously. From Davis's double-double to a double-kill. History tells us one thing: in the commodities market, long-term contracts are never a one-sided &#34;floor.&#34; They protect buyers in an upward cycle and sellers in a downward cycle, but only if both sides have the ability and willingness to fulfill their obligations. The moment when long-term contracts are truly needed is precisely the moment when they are most likely to fail. This isn't to say that Micron is definitely a bubble. AI's demand for computing power and storage may indeed be structural, LTAs may truly rewrite industry rules, and a trillion-dollar market capitalization may just be the beginning. But when the entire Wall Street is simultaneously shouting &#34;This time is different,&#34; it's 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 a bubble. However, it took Cisco approximately 25 years, until today's AI era, to surpass its closing high during the dot-com bubble, even though the internet has indeed changed everything.</div></div></div>

SanDisk hiked NAND pricing by 50% in one move on surging AI demand, with the supply crunch drawing 'windfall profit' comparisons and forcing some downstream buyers to suspend quoting. The news sent storage names sharply higher — $SNDK surged 10%+ intraday, with Micron, Western Digital, and Seagate all rallying in tandem. Bank of America raised its PT on SanDisk to $2,100. Is this storage rally still worth chasing?

SanDisk hiked NAND pricing by 50% in one move on surging AI demand, with the supply crunch drawing 'windfall profit' comparisons and forcing some downstream buyers to suspend quoting. The news sent storage names sharply higher — $SNDK surged 10%+ intraday, with Micron, Western Digital, and Seagate all rallying in tandem. Bank of America raised its PT on SanDisk to $2,100. Is this storage rally still worth chasing?

SanDisk Hikes Prices 50% — Is the Memory Sector Set to Surge?

XIACY

SOXX

SOXQ

FTXL

SOXL

DRAM

Xiaomi's Q1 2026 report shows a 10.9% revenue drop to 99.1 billion yuan and a 43.1% decline in net profit. The surge in memory prices, nearly fourfold, is impacting costs, but Xiaomi won't pass these to consumers. Meanwhile, Micron Technology's stock soared over 19%, with UBS raising its target price significantly. Analysts suggest that Long-Term Agreements (LTAs) may stabilize the memory market, driven by AI demand from major tech firms. Goldman Sachs, however, has been cautious, maintaining a neutral rating despite the stock's rise.

- Xiaomi reported Q1 2026 revenue of 99.1 billion yuan, a 10.9% year-on-year decline, attributing struggles to increased memory costs and reduced phone model offerings.  
- In contrast, Micron Technology's stock surged over 19%, driven by bullish target price revisions from multiple brokerages, reflecting strong sentiment around long-term contracts securing future supply amid an AI-driven demand boom.  
- Analysts caution that reliance on high prices and contracts could expose Micron to significant risk if AI demand slows, reminiscent of past market cycles where optimism led to substantial losses.

The trillion-dollar frenzy of selling memory modules has been halved in the profits of buying them.