Longxin Technology's IPO and valuation range from 3.2 to 7.76 trillion CNY

Original author: Long Yue

Source: Wall Street Journal

ChangXin Memory Technologies (688825) is set to list today, becoming the largest IPO on the STAR Market. The IPO offering price is RMB 8.66, with a total post-offering share capital of 66.881 billion shares (before exercise of the over-allotment option), resulting in a total market capitalization of RMB 579.188 billion. However, the market clearly has no intention of staying at this price.

Analyst Li Jiu from Northeast Securities conducted a valuation of ChangXin from three independent perspectives in a research report, concluding with a range of RMB 3.2 to 5.7 trillion. On the same day, Nomura Securities initiated coverage with a "Buy" rating and a target price of RMB 116, implying a 1,239% upside and a market capitalization of approximately RMB 7.76 trillion—1.4 times the upper limit of Northeast Securities’ estimate. The core divergence between the two institutions lies in their assessment of ChangXin’s long-term market share ceiling—Northeast Securities’ base case assumes 17%, while Nomura bets on greater market share potential and a higher growth premium.

The above valuation may not be exaggerated. ChangXin is a unique entity in the A-share market: the leading pure-play DRAM IDM with full end-to-end capabilities in design and manufacturing, currently in a phase of explosive performance growth driven by a cyclical turnaround and market share expansion. Its product portfolio covers DDR4/5 and LPDDR4X/5/5X, and its products have been adopted by Alibaba, Tencent, ByteDance, and major smartphone supply chains. According to Omdia data, the company held a global market share of 7.67% in Q4 2025, ranking first in China and fourth globally. Benefiting from rising memory prices and increased volume of high-end products, the company’s earnings elasticity is accelerating.

The DRAM supply-demand gap persists, and ChangXin benefits from dual advantages of rising volume and prices alongside domestic substitution. The real question isn't whether it's valuable, but which yardstick to use to measure it.

Logic: Since DRAM is a global unified market, the market capitalization of overseas-listed memory manufacturers already incorporates pricing for every percentage point of market share. Using the market capitalization of comparable U.S. stocks, derive the value per 1% of future global market share, then multiply by CXMT’s projected market share.

Action: Micron and SanDisk each hold exactly 13% of the NAND market share; therefore, the difference between Micron’s market capitalization (a company with both DRAM and NAND businesses) and SanDisk’s market capitalization (a pure NAND company) equals the market value attributable to Micron’s DRAM business—$1022 billion minus $230.8 billion equals $791.2 billion. Dividing this by Micron’s forward DRAM market share of 19.85% yields approximately $39.86 billion per 1% of forward DRAM market share.

Conclusion: As a pure-play DRAM company, ChangXin is projected to have a future market share of 17% (currently around 8%), corresponding to a market capitalization of approximately $677.676 billion, or about RMB 4.58 trillion (at an exchange rate of 6.77). After excluding the minority interest share of approximately 24%, the parent company's attributable value is approximately RMB 3.49 trillion.

Backtesting verification: Hynix's calculated market cap is 9.44% higher than actual, while Kioxia is only 0.66% higher, closely matching the actual market cap.

The second approach is more fundamental: it directly forecasts ChangXin’s own profits without relying on external benchmarks. The cost structure of memory fabs is highly standardized, with fixed costs primarily consisting of depreciation, determined by the scale of capital expenditures; variable costs change linearly with shipment volume. Since the prospectus does not disclose actual wafer capacity data, fixed asset value is used as a proxy for capacity, multiplied by utilization rate and sales-to-production ratio to estimate sales volume, and then combined with ASP to calculate revenue.

Logic: Break down revenue (capacity × utilization rate × sales-to-production ratio × ASP) and costs (fixed cost depreciation + variable costs), forecast net profit, then apply a P/E multiple.

Key prediction:

• Revenue of RMB 471.6 billion in 2027, gross margin of 86.96%, net profit of RMB 374.7 billion (consolidated basis)
• Excluding minority interest (assuming a constant share of 24%), net profit attributable to parent shareholders is approximately 284.8 billion yuan.

Valuation: Northeast Securities believes that Micron and SK Hynix have P/E ratios of 7.51x and 7.94x for 2027, respectively, but ChangXin is in a phase of rapid market share growth (with projected long-term share reaching around 30%), warranting a growth premium; applying a P/E range of 10x to 15x and excluding non-controlling interests, the market capitalization corresponding to net profit attributable to parent company shareholders is approximately RMB 2.85 trillion to RMB 4.27 trillion.

Logic: Divide the market capitalization of overseas memory manufacturers by their monthly production capacity to obtain "market capitalization per ten thousand units per month," then multiply by CXMT's planned production capacity.

Reference: The market capitalization corresponding to the monthly production capacity of 10,000 wafers for the three major manufacturers is concentrated in the range of $15.8–19.8 billion—Hynix at $16.045 billion, Micron at $19.78 billion, and Samsung at $15.891 billion.

Conclusion: ChangXin's production capacity in 2027 will be 450,000 wafers/month, corresponding to a market capitalization:

• Optimistic scenario (average of three major manufacturers: $17.2 billion per 10,000 units): 5.2518 trillion yuan
• Base case (including average of Taiwanese manufacturers: $13.9 billion per wafer): 4.2327 trillion yuan

After excluding the net profit attributable to non-controlling interests, it amounts to approximately RMB 3.22 trillion to 3.99 trillion.

Northeast Securities noted that non-controlling interest in ChangXin's 2025 earnings accounts for as high as 73.76%, far exceeding Samsung, SK Hynix, and Micron (all under 1%), and the valuation must exclude this impact.

Assuming the minority interest share remains at 24% in 2026 and 2027, the three methods yield the following conclusion: the fair valuation, after excluding the impact of minority interest, is between 3.2 and 5.7 trillion.

The three perspectives use different data and logical chains, but their final consolidated figures all converge near the 3 to 4.3 trillion range. This convergence itself is a signal: under the current market share and capacity assumptions, this pricing level exhibits strong internal consistency.

On July 27, Nomura Securities also initiated coverage of CXMT with a more aggressive assessment.

The stock is initiated with a "Buy" rating and a target price of RMB 116, representing approximately 20x P/E—twice the current valuation of Micron (around 10x) and more than twice that of SK Hynix.

Based on an IPO offering price of 8.66 yuan, the target price of 116 yuan implies an implied upside of 1,239.5%, corresponding to a market capitalization of approximately RMB 7.76 trillion.

This figure far exceeds the upper bound of Dongbei Securities' valuation range of 5.7 trillion, and the approximately 2 trillion difference between the two essentially reflects differing bets on two key variables: where Longxin's market share ceiling lies, and how much growth premium the market should assign to this company.

Nomura's rationale for the 20x P/E premium is based on three key judgments.

First, structural supply constraints will persist for several years. The bank’s core argument is: “Global storage supply is unlikely to ease over the next few years.” Samsung, SK Hynix, and Micron have significantly shifted their capital expenditures toward HBM and advanced processes, structurally suppressing new supply in generic DRAM. This means the generic DRAM market, where ChangXin operates, will maintain a supply-demand imbalance for an extended period, rather than following the traditional storage cycle of “two years up, two years down.”

Second, the logic behind ChangXin’s market share growth is “accelerating,” not linear. The firm believes that as ChangXin continues to expand its production capacity and transitions from fourth-generation to fifth-generation processes, its rate of market share gain in the global general-purpose DRAM market will exceed market expectations. Given its current share of approximately 8%, the long-term potential implied in Nomura’s model clearly far exceeds the 17% upper limit assumed by Northeast Securities. Based on Nomura’s implied long-term share assumption and a target market capitalization of NT$7.76 trillion, the potential market share range could be between 25% and 30%, or even higher.

Third, the combination of domestic substitution and AI demand provides a dual growth premium. The firm believes that CXMT is not only a cyclical storage play but also a thematic play on “domestic substitution.” Chinese cloud and smartphone manufacturers’ willingness to procure domestic DRAM continues to rise, providing CXMT with additional growth independent of global cycles. Meanwhile, demand for DRAM from AI servers is growing exponentially, with server DRAM capacity per unit nearly 80 times that of smartphones. This shift in demand structure will sustain a long-term upward pressure on ASPs. Combining these two drivers, Nomura believes CXMT should command a valuation premium over its overseas peers, rather than a discount.

In other words, Nomura does not view 2026 as the peak, but rather as a starting point.

In terms of specific financial forecasts, Nomura predicts that ChangXin's revenue and net profit attributable to shareholders will increase by 63% and 74%, respectively. Key drivers include capacity expansion from 270,000 wafers per month in 2025 to 450,000 wafers per month in 2027, higher wafer value due to process migration, and continued upward pressure on DRAM average selling prices amid tighter supply. Nomura’s profit forecasts are more aggressive in absolute terms than those of Northeast Securities, and the 20x P/E multiple assumption further amplifies the final valuation outcome.

The more aggressive hypothesis is supported by supply and demand data: global general-purpose DRAM capacity estimates show a shortfall still exists in 2027; DRAM contract prices in Q1 2026 surged环比 by 93% to 98%, significantly exceeding prior expectations; CXMT’s gross margin in Q1 2026 has risen to 79.16%, with a net profit attributable to parent shareholders of RMB 24.762 billion for the quarter.

The height of the price increase cycle is currently adjusting the input assumptions of all models.

In 2019, CXMT (formerly RuiLi Integrated) launched China's first domestically produced and mass-produced 8Gb DDR4 chip, marking a breakthrough in achieving domestic DRAM production from zero to one.

Seven years later, the company has become the largest DRAM manufacturer in China and the fourth-largest globally. According to Omdia data, CXMT’s global market share reached 7.67% in Q4 2025.

On the product front, CXMT has covered all generations of DDR4/5 and LPDDR4X/5/5X, and ceased its own DDR4 production by the end of 2024, shifting all capacity toward higher-value products such as DDR5 and LPDDR5/5X. Its customers include Alibaba, Tencent, ByteDance, and major smartphone supply chains.

In terms of production capacity, the company operates three 12-inch wafer fabs—two in Hefei and one in Beijing. Northeast Securities expects capacity to increase from 270,000 wafers per month in 2025 to 450,000 wafers per month in 2027, raising its global market share from 14% to 17%.

Changxin's financial path follows the classic playbook of a capital-intensive memory factory—fixed costs are front-loaded during downturns, leading to losses, while profits are rapidly unlocked during upturns.

The key nodes are as follows:

• 2025: Net profit attributable to shareholders turned from a loss of RMB 16.34 billion to a profit of RMB 1.875 billion, with the consolidated gross profit margin rising to 40.99%, nearly matching Samsung's (39.38%).
• Q1 2026: Quarterly revenue of RMB 50.8 billion (+719% YoY), gross margin of 79.16%, net profit attributable to parent company of RMB 24.762 billion
• First half of 2026: Management expects revenue of RMB 110-120 billion and net profit attributable to parent company shareholders of RMB 50-57 billion.

The core driver of profit surge is price. According to TrendForce's latest survey in June 2026, the contract price for general-purpose DRAM in Q1 2026 surged approximately 93%-98% quarter-over-quarter, far exceeding the previously predicted double-digit range.

Northeast Securities also lists four major risks:

1. Demand fell short of expectations: AI server construction is slowing down or consumer electronics recovery is weak.
2. Periodic price decline: In 2022-2023, prices experienced a significant drop of up to 50% compared to the previous cycle peak.
3. Production capacity and technological iteration lagging behind expectations: Delays in the development of the fifth-generation process platform will impact volume and price realization.
4. International trade friction and supply chain constraints: Escalating geopolitical tensions may exacerbate instability in the industrial chain.