ChangXin Technology’s July 27 listing quickly turned into a broad market exercise in pricing the future of the memory industry. The stock opened at RMB 49.50, fell to RMB 38.11, climbed to RMB 55.03, and finished the day back near RMB 49.
That same day, investors in Shanghai, Seoul and New York were all working through different versions of the same question. In China, the Shanghai Composite rose 1.15%, the Shenzhen Component gained 2.72%, the STAR 50 added 1.16%, and the semiconductor materials and equipment index climbed 4.21%. South Korea’s benchmark took a more jagged V-shaped path. U.S. stocks opened higher and then gave back most of the advance, with chip names reversing more sharply; the Philadelphia Semiconductor Index went from a 1.1% intraday gain to a 2.23% loss at the close.
The article calls it a “super storage week.” ChangXin went public on Monday and faced public-market price discovery. On Wednesday and Thursday, SK Hynix and Samsung Electronics were due to report second-quarter results, providing fresh numbers on profit, inventory and capital spending. At the same time, global investors were still dealing with U.S. long-term Treasury yields above 5%.
Heavy turnover marked the first trading session
Trading in ChangXin reflected sharply different time horizons. Some investors sold immediately after receiving IPO shares. Others added to positions during the drop. The article describes those reactions as neither fully rational nor irrational, but all of them appeared in the same candlestick, with a long upper shadow and a long lower shadow.
At around 9:44 a.m. on July 27, the stock touched RMB 38.11, its intraday low. Just over 10 minutes after the open, about RMB 65 billion had already changed hands. Mid-sized and small orders posted a combined net outflow of about RMB 37.8 billion, while extra-large and large orders absorbed supply in nearly equal size.
A turnover rate of 66.4% meant that, out of 4.503 billion shares available for trading on the first day, roughly 2.99 billion moved from original allottees to new buyers. Millions of winning IPO shares were cashed out at once, sell orders stepped through the bid, and only when the price neared RMB 38 did large pools of capital begin to absorb enough supply to push the stock back above RMB 46 within minutes.
The article argues that RMB 38.11 was not the output of a valuation model. It was closer to the price left behind when two clocks collided: one focused on locking in multiple times the book profit from the IPO allocation, the other willing to pay for the future of China’s DRAM industry.
What did RMB 49 actually buy?
The report frames the central issue this way: did RMB 49 reflect ChangXin’s current earnings, the strength of the DRAM cycle, or the possibility of China’s storage industry over the next decade?
On Monday, the A-share market offered a first answer: RMB 49 a share, implying a market capitalization of about RMB 3.28 trillion. But the article says that was only an opening quote, not a final conclusion.
In its view, ChangXin’s closing market value was about RMB 2.1 trillion above the midpoint of a fundamentals-based model. Investors did not wait for HBM, yield improvement, capacity expansion and profit delivery to show up one by one. Part of that industrial future was written into the stock price right away.
That leaves the valuation debate hanging on execution. Whether RMB 49 is expensive depends on whether ChangXin can gradually turn more than RMB 2 trillion in what the article calls “option value” into capacity, technology, market share and profits that can survive through the cycle.
A veteran private equity investor quoted in the piece put it this way: “ChangXin Technology is the first truly world-class technology company in the A-share market and the valuation anchor for large-cap technology in A-shares. Its next move will still far exceed expectations. Nomura has given ChangXin a long-term target price of RMB 112. Goldman Sachs mentioned on a conference call that ChangXin’s HBM progress may be much better than expected. This is a new shift in valuation logic.” He said he had tried to add to his position around RMB 40 but was unable to buy because of network issues in the trading system.
SK Hynix and Samsung results became reference points
The article says the next pricing clues would come from South Korea. SK Hynix’s earnings were expected to show how much money AI memory can generate. Samsung Electronics’ report, in turn, was expected to offer a market answer on how scale, technology and capital spending should be valued.
Looking at profit alone, the market had little reason to be disappointed. Samsung had already guided for about KRW 171 trillion in second-quarter sales and about KRW 89.4 trillion in operating profit, up more than 18 times from a year earlier. South Korea’s semiconductor exports for the first 20 days of July rose 180.6% year over year, giving the demand side a concrete data point.
But the article says investors were no longer satisfied with the simple question of how much companies earned in the second quarter. What they wanted next included third-quarter contract prices for DRAM and NAND, customer qualification and yields for HBM4, whether inventories were moving back into a restocking phase, how much new capacity would arrive in 2027 and 2028, and whether capital spending could ultimately become free cash flow.
ChangXin was being written into that same global supply table. Inside China, more capacity means progress in import substitution. In the global market, it means that a fourth DRAM maker is reaching scaled output alongside Samsung Electronics, SK Hynix and Micron Technology. The stronger demand becomes, the stronger the incentive to expand. The faster the expansion, the earlier investors begin worrying about the next supply wave. That, the article says, is the hard part of cyclical industries: today’s shortage often plants the seeds of tomorrow’s surplus.
Three clocks were pricing the sector at once
The report breaks the story into three timelines.
- Industrial decisions run on a yearly clock: orders, fabs, technology iteration and customer qualification all take time.
- Financial statements run on a quarterly clock: revenue, gross margin, depreciation and capital spending settle quarter by quarter.
- Capital runs on a daily clock: long-bond yields, fund redemptions and foreign flows are repriced every day.
The intraday reversal in U.S. storage stocks on July 27 raised another question. Is ChangXin simply a beneficiary of the global upcycle, or is it already a participant that could change the future supply structure?
In less than 10 years, according to the article, ChangXin has grown into the fourth scaled producer in the global DRAM market, and its market capitalization has entered the ranks of semiconductor heavyweights. Even so, gaps remain with Samsung Electronics, SK Hynix and Micron in earnings scale, advanced products, customer mix and technology ecosystem.
From 8Gb DDR4 to the top tier of A-shares
The article retraces part of that rise. ChangXin appeared at the World Manufacturing Convention in September 2019 with its 8Gb DDR4 product, and received its first order in November that year. In less than seven years, the company rose to become the largest stock by market capitalization in the A-share market, according to the report.
It also cites remarks from 2022, when ChangXin chairman and CEO Zhu Yiming spoke at the World Integrated Circuit Conference and the second IC China Expo in a presentation titled “Challenges to the Globalized Supply Chain of Integrated Circuits Under the New Situation.” Zhu said globalization of the industrial chain was the course of history and that cooperation was the only path to mutual benefit.
Public-market pricing, however, brings a different set of tests. The company still has to get through swings in sentiment, profit delivery, the cycle, technology and global competition.
Repricing happened on both sides of the Pacific
The article says July 27, 2026 belonged to the STAR Market, seven years after its launch. Peng Yigang, deputy director of the market development department at the Shanghai Stock Exchange, said at a World Artificial Intelligence Conference forum that companies are stars, and the STAR Market itself is also a new star in China’s capital market. He invited attendees to “join hands and head toward the sea of stars.”
By the close on July 27, the book value of Zhu Yiming’s shareholding exceeded RMB 78 billion. Across the Pacific, Micron, SanDisk and a group of semiconductor stocks were sold off. The possibility of Chinese memory-capacity expansion and domestic immersion DUV lithography tools was quickly written into the pricing of global chip stocks.
The article’s conclusion is not framed as a final verdict, but as a split in how markets responded. In China, investors added a premium for ChangXin’s scarcity. Overseas, investors began assigning a fresh discount for the competition it might bring. On the evening of July 27, the more pressing issue for Zhu may not have been the RMB 49 closing price, but whether the RMB 57.9 billion in fundraising, or as much as RMB 66.6 billion if the greenshoe were fully exercised, could be turned into equipment, yield, capacity and next-generation products — and whether the future paid for in advance by the market could be delivered quarter by quarter in later earnings reports.
The article was originally published by the WeChat public account of Economic Observer and written by Ouyang Xiaohong.