Where to begin with CXMT's valuation?

CXMT is set to debut on the STAR Market today, marking the largest IPO on the board. The IPO price is set at 8.66 yuan, which, based on conservative estimates, implies a market capitalization of around 520 billion yuan. However, the market does not appear inclined to linger at this price point.

Analyst Li Jiu from Northeast Securities has provided three independent valuation perspectives for CXMT, with conclusions converging in the 3.2 to 5.7 trillion yuan range. On the same day, Nomura initiated coverage with a "Buy" rating and a target price of 116 yuan, implying a potential upside of 1,239% and a market cap of approximately 7.76 trillion yuan—1.4 times the upper limit of Northeast Securities' estimate. The core disagreement between the two institutions lies in their assessment of CXMT's long-term market share ceiling. Northeast Securities assumes a baseline of 17%, while Nomura bets on a larger market share and a higher growth premium.

These valuation figures may not be exaggerated. CXMT is a unique entity in the A-share market: a pure DRAM IDM leader with full capabilities across design and manufacturing. It is currently in a period of explosive earnings growth driven by a "cycle reversal plus market share increase." The company's products cover DDR4/5 and LPDDR4X/5/5X, and they have entered the supply chains of Alibaba, Tencent, ByteDance, and major smartphone manufacturers. According to Omdia data, CXMT held a 7.67% global market share in the fourth quarter of 2025, ranking first in China and fourth globally. Benefiting from memory price increases and the volume ramp-up of high-end products, the company's earnings elasticity is accelerating.

The DRAM supply-demand gap persists, and CXMT enjoys the dual benefits of "volume growth and price increases" combined with "domestic substitution." The real question is not whether the company is valuable, but which valuation metric is most appropriate.

Approach One: Relative Valuation by Market Share – Target Market Cap of Approximately 3.49 Trillion Yuan

The logic is that since DRAM is a globally unified market, the market capitalizations of overseas-listed memory companies already reflect the market's pricing for each percentage point of market share. By using the market caps of comparable US-listed companies, one can reverse-engineer the value attributed to "every 1% of future global market share" and then multiply it by CXMT's projected future share.

The methodology involves Micron Technology and Sandisk, which have equal NAND market shares (both at 13%). Therefore, subtracting Sandisk's market cap (a pure NAND company) from Micron's market cap (a DRAM and NAND company) isolates the value of Micron's DRAM business: $1.022 trillion - $230.8 billion = $791.2 billion. Dividing this by Micron's 19.85% future DRAM market share yields a value of approximately $39.86 billion for every 1% of future DRAM market share.

As a pure DRAM company with a projected future market share of 17% (currently around 8%), CXMT would have a corresponding market cap of about $677.68 billion, or roughly 4.58 trillion yuan (at an exchange rate of 6.77). After deducting approximately 24% for minority interests, the attributable market cap would be around 3.49 trillion yuan.

Back-testing shows that this method yields a value for SK Hynix that is only 9.44% higher than its actual market cap, and for Kioxia, it is only 0.66% higher, indicating a reasonable degree of accuracy.

Approach Two: Earnings-Based PE Valuation – Target Market Cap of 2.85 Trillion to 4.27 Trillion Yuan

The second method is more fundamental, forecasting CXMT's own profits without relying on external benchmarks. The cost structure of memory fabs is highly standardized. Fixed costs are primarily depreciation, determined by capital expenditure, while variable costs scale linearly with output. Since the prospectus does not disclose actual wafer capacity data, the calculation uses fixed assets as a proxy for capacity, multiplies it by utilization and sales rates to estimate sales volume, and then multiplies that by the average selling price (ASP) to derive revenue.

The logic involves breaking down revenue (capacity × utilization rate × sales rate × ASP) and costs (fixed depreciation + variable costs) to forecast net profit, which is then valued using a PE multiple.

Key predictions include: revenue of 471.6 billion yuan in 2027, a gross margin of 86.96%, and net profit of 374.7 billion yuan (on a total basis). After deducting minority interests (assumed to remain at 24%), attributable net profit would be approximately 284.8 billion yuan. Northeast Securities notes that Micron and SK Hynix trade at 7.51x and 7.94x their 2027 earnings, respectively. However, given CXMT's rapid market share growth phase (with a projected future share of around 30%), they apply a growth premium and use a 10-15x PE multiple. After deducting minority interests, this results in an attributable market cap of 2.85 trillion to 4.27 trillion yuan.

Approach Three: Relative Valuation by Unit Capacity – Target Market Cap of 3.22 Trillion to 3.99 Trillion Yuan

The logic here is to divide the market capitalization of overseas memory fabs by their monthly capacity to derive a "value per 10,000 wafers/month of capacity," which is then multiplied by CXMT's capacity plan.

The reference point is that the value per 10,000 wafers/month for the three major memory manufacturers ranges from $15.8 billion to $19.8 billion: SK Hynix at $16.045 billion, Micron at $19.78 billion, and Samsung at $15.891 billion.

With CXMT's capacity projected at 450,000 wafers per month in 2027, the corresponding market cap is: in an optimistic scenario (using the average of the three majors at $17.2 billion/10k wafers) it would be 5.2518 trillion yuan; in a neutral scenario (including Taiwanese fabs, average $13.9 billion/10k wafers) it would be 4.2327 trillion yuan. After deducting minority interests, this yields a range of 3.22 trillion to 3.99 trillion yuan.

Summary of the Three Methods: Convergence at 3.2 Trillion to 5.7 Trillion Yuan

Northeast Securities points out that CXMT's minority interest ratio was as high as 73.76% in 2025, far exceeding that of Samsung, SK Hynix, and Micron (all below 1%), so any valuation must exclude this impact. Assuming the minority interest ratio remains at 24% in 2026 and 2027, the three methods yield a reasonable valuation range of 3.2 to 5.7 trillion yuan after deducting the minority interest impact.

The three approaches use different data and logic chains, but the final attributable market cap ranges all converge around 3 to 4.3 trillion yuan. This convergence itself is a signal: under the current assumptions for market share and capacity, this level of pricing is highly self-consistent.

The real point of divergence is "how high the future market share can go." Northeast Securities assumes a baseline of 17%, while Nomura's implied assumption is more aggressive. The 116 yuan target price, corresponding to a market cap of about 7.76 trillion yuan, suggests the market is willing to pay for a larger market share and a higher premium.

Supporting this more aggressive assumption are supply-demand data: global general DRAM capacity calculations show a persistent supply gap in 2027; general DRAM contract prices surged 93% to 98% quarter-over-quarter in the first quarter of 2026, far exceeding previous double-digit forecasts; and CXMT's gross margin rose to 79.16% in Q1 2026, with a single-quarter attributable net profit of 24.762 billion yuan. The height of the price upcycle is constantly revising the input assumptions of all models.

Nomura's Perspective: Target Price of 116 Yuan, Implying 1,239% Upside

On July 27, Nomura also initiated coverage of CXMT with a more aggressive outlook. The firm started with a "Buy" rating and a target price of 116 yuan, corresponding to a PE of about 20x—twice the current valuation of Micron (about 10x) and more than double that of SK Hynix. Based on the IPO price of 8.66 yuan, the 116 yuan target implies a potential upside of 1,239.5% and a market cap of approximately 7.76 trillion yuan.

This figure far exceeds the upper limit of Northeast Securities' valuation range of 5.7 trillion yuan. The roughly 2 trillion yuan gap between them is essentially a bet on two core variables: the ceiling of CXMT's market share and the growth premium the market should assign to the company.

2026 is Just the Beginning! Domestic Substitution and AI Demand Drive Double Growth Premium, Says Nomura

Nomura's logic for assigning a 20x PE premium is built on three key judgments.

First, structural supply tightening will persist for several years. The firm's core argument is that "global memory supply is unlikely to be loose in the coming years." Capital expenditure by Samsung, SK Hynix, and Micron has shifted massively toward HBM and advanced processes, structurally suppressing new supply for general-purpose DRAM. This means CXMT's general DRAM market will likely remain undersupplied for a considerable time, rather than following the traditional memory cycle of "two years up, two years down."

Second, CXMT's market share growth is "accelerating" rather than "linear." Nomura believes that as CXMT's capacity expands and its process technology migrates from the fourth to the fifth generation, its global market share in general DRAM will increase faster than market expectations. The current share of about 8%, in Nomura's model, clearly implies a future share well beyond the 17% ceiling assumed by Northeast Securities. Reverse-engineering from Nomura's 7.76 trillion yuan target market cap suggests an implied future market share in the 25% to 30% range, or even higher.

Third, domestic substitution combined with AI demand provides a double growth premium. The firm believes CXMT is not just a memory cycle play but also a "domestic substitution" theme stock. The increasing willingness of Chinese cloud and smartphone companies to purchase domestic DRAM provides CXMT with additional incremental demand independent of the global cycle. Concurrently, AI servers' demand for DRAM is growing exponentially, with server DRAM content per unit being nearly 80 times that of a smartphone. This structural shift in demand will support a long-term rise in the ASP floor. With these two factors combined, Nomura believes CXMT should command a higher valuation premium than its overseas peers, not a discount.

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

In its specific financial forecasts, Nomura predicts CXMT's sales and attributable net profit will grow 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, increased value per wafer from process migration, and continued upward DRAM prices amid tight supply. Nomura's absolute earnings forecasts are more aggressive than Northeast Securities', and the 20x PE multiple further amplifies the final valuation result.

From Zero to Global Number Four: CXMT's Seven-Year Journey

In 2019, CXMT (formerly known as Rui Li Integrated) launched China's first independently mass-produced 8Gb DDR4, marking a breakthrough from zero to one for domestic DRAM. Seven years later, the company is now the largest DRAM manufacturer in China and the fourth largest globally. According to Omdia, CXMT held a 7.67% global market share in Q4 2025.

Its product line now covers all generations of DDR4/5 and LPDDR4X/5/5X. By the end of 2024, it stopped production of its own DDR4, shifting all capacity to high-value products like DDR5 and LPDDR5/5X. Its customer base includes Alibaba, Tencent, ByteDance, and major smartphone supply chains.

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

Financial Inflection Point: Stunning Profit Elasticity in a Price Upcycle

CXMT's financial trajectory follows the classic script for a capital-intensive memory fab—fixed costs are incurred upfront during a downcycle, leading to losses, while profits are released rapidly during an upcycle. Key milestones include: in 2025, attributable net profit swung from a loss of 16.34 billion yuan to a profit of 1.875 billion yuan, with a comprehensive gross margin of 40.99%, roughly on par with Samsung's 39.38%. In Q1 2026, single-quarter revenue reached 50.8 billion yuan (up 719% year-over-year), with a gross margin of 79.16% and attributable net profit of 24.762 billion yuan. Management expects H1 2026 revenue to be 110-120 billion yuan and attributable net profit to be 50-57 billion yuan. The core driver of this profit explosion is price. According to TrendForce's June 2026 survey, general DRAM contract prices surged approximately 93% to 98% quarter-over-quarter in Q1 2026, far exceeding previous double-digit forecasts.

Northeast Securities also lists four key risks: lower-than-expected demand from a slowdown in AI server construction or weak consumer electronics recovery; a cyclical price downturn, as seen in 2022-2023 when prices fell by up to 50% from cycle peaks; slower-than-expected capacity and technology iteration, which could hinder volume and price realization if fifth-generation process platform development lags; and international trade friction and supply chain constraints, as geopolitical tensions could increase industry instability.