Quick ReadPlain-language overview · read this first
NSIG is one of China’s leading semiconductor silicon wafer companies. The report’s stance is Watch: do not rush to buy, because the current price leaves no cheap margin.
What does it mainly do? Chipmaking relies on a foundational material called silicon wafers. NSIG develops and manufactures these wafers, then sells them to chipmakers such as TSMC and SMIC. This business is hard for outsiders to enter: the world’s top five suppliers account for more than 90% of the market, and newcomers need not only the technology but also long customer validation before they can supply at scale. That gives NSIG meaningful standing in China.
Still, this is not an easy business. In 2025, its shipments reached a new high since listing, but both of its two main products were sold at a loss per unit, with negative gross margins. In other words, scale kept expanding, yet profitability had not followed. This is the report’s main concern: the company is spending heavily on capacity expansion while cash keeps flowing out of the business. Operating cash flow has been negative for two consecutive years. When cash is insufficient, the gap is filled with debt and new share issuance; as the share count rises, existing shareholders’ stake gets diluted.
Is it expensive to buy now? Its total market value is about 87.4 billion yuan. The report calculates that this price equals 23.5 times one year of revenue, more expensive than global leaders, while its earnings and cash flow are worse. The report estimates a reasonable price of roughly 7 to 11 yuan per share, while the current share price is already above 26 yuan, meaning the good news of the next ten years has effectively been priced in upfront. The biggest risk to watch is this: if shipments rise but pricing and costs fail to keep up, domestic substitution may ultimately deliver only revenue, not shareholder returns, which could lead to a long-term loss of principal.
The report’s final stance is: the direction matters and deserves tracking, but it is not currently suitable as a long-term value investment. A more prudent approach is to put it on the watchlist and wait for a lower price or clearer signs of improvement.
The above is only a plain-language explanation of this report, not investment advice. The stock market involves risk; invest with caution.
LeadNational Silicon Industry Group is the mainland China leader in domestic substitution for semiconductor wafers, with core products spanning 300mm, 200mm and smaller polished wafers, epitaxial wafers, and SOI wafers for domestic fabs. Revenue reached a post-listing high of about RMB 3.7 billion in 2025, but gross margins in both the 300mm and 200mm wafer businesses remained negative, the company posted losses for two straight years, operating cash flow was RMB -559 million, and it remains in a phase of capacity ramp-up and repeated equity financing. Research rating Watch: an important strategic materials company worth tracking, but not yet a mature cash-flow asset with a compelling margin of safety.
Prices in the article are as of publication; see the valuation band above for the live price.
Conclusion First
Initial Conclusion
| Item | Conclusion |
|---|---|
| Investment rating | Watch |
| Does the current price offer a margin of safety? | No |
| Better-suited investors | Industrial-trend / domestic-substitution investors; less suitable for balanced, conservative long-term value investors |
| Business understandability | 4/5 |
| Industry attractiveness | 3/5 |
| Moat strength | 3/5 |
| Management and capital allocation | 2.5/5 |
As of 2026-06-09, Reuters delayed quotes showed National Silicon Industry Group trading at about RMB 26.44 per share, with roughly 3.305 billion shares outstanding. On that basis, its market capitalization was about RMB 87.38 billion. Against 2025 revenue of RMB 3.716 billion and Q1 2026 net assets attributable to the parent of RMB 16.835 billion, the current valuation is roughly 23.5x price-to-sales and 5.2x price-to-book. That is not cheap for a heavy-asset materials company still ramping capacity, after huge losses in both 2024 and 2025 and deeply negative free cash flow.
My core judgment has four parts. First, the business is understandable: it essentially sells semiconductor wafers that require long qualification cycles; customers are mainly fabs, and the product is important, technically difficult, and slow to validate. Second, this has not yet proven itself to be a “good business”: revenue reached a post-listing high in 2025, but gross margins in both the 300mm and 200mm-and-below wafer businesses remained negative, showing that scale expansion has not yet translated into returns. Third, the company has industrial standing and technical barriers, but it has not shown the high returns, strong cash flow, and pricing power that Buffett tends to prefer. Fourth, buying at the current price is closer to betting that “domestic substitution succeeds over the next decade + capacity ramps smoothly + gross margins recover” all happen together, rather than buying an already mature cash-flow asset.
The largest uncertainties are highly concentrated. First, whether 300mm capacity expansion can truly turn into sustainable positive gross margins and positive free cash flow; second, when the 200mm-and-below business can emerge from pricing and demand pressure; third, whether the company will continue to rely on equity or debt financing, diluting per-share value. In 2025, the current ratio fell to 1.53, the debt-to-asset ratio rose to 40.33%, and interest coverage dropped to -6.98, indicating that financing pressure has already increased versus the past.
One-Sentence Conclusion
National Silicon Industry Group is a strategic materials company that is important, difficult to build, and worth tracking, but it is not yet a long-term value investment target whose cash flow has been validated and whose current valuation is cheap.
Business Understanding and Industry Structure
How This Company Actually Makes Money
National Silicon Industry Group's core business is not complicated: it develops, manufactures, and sells semiconductor wafers. Its core products include 300mm semiconductor wafers, 200mm and smaller polished / epitaxial wafers, and SOI wafers, which are ultimately used in chips for memory, image processing, general-purpose processors, power devices, sensors, RF, analog, and discrete devices. The company earns revenue by selling products directly to downstream fabs. Its 2022 annual report summary explicitly stated that its profit model is to “sell semiconductor wafers to downstream chip manufacturers to generate revenue and profit,” with direct sales as the main sales model.
At the customer level, this is not a consumer-facing brand business. It is a “qualified materials supply business” serving large customers. Disclosed customers include TSMC, UMC, GlobalFoundries, STMicroelectronics, TowerJazz, SMIC, Hua Hong Grace, Huali Microelectronics, and China Resources Microelectronics, among other domestic and overseas fabs. By 2025, the company further disclosed that Shanghai Xinsheng's 300mm business had achieved full coverage across logic, memory, image sensor, and power applications, as well as full coverage of major domestic customers. This means revenue durability depends more on customer qualification, yield, delivery, and cost than on brand premium.
The “recurring” nature of revenue is weaker than it may appear. Once qualification is complete, it is indeed difficult for customers to switch suppliers, and supply relationships can last for years. On the other hand, the industry is affected by fab utilization, industry inventory cycles, product pricing, and the company's own yield ramp. In Q1 2026, revenue increased 35.22% year over year, mainly because 300mm wafer sales volume grew by more than 90% year over year. But profit did not improve in tandem because prices were under pressure, fixed costs rose, R&D investment increased substantially, and foreign-exchange and interest expenses increased. In other words, this is not a business where “the more it sells, the more stable it becomes”; it is a business where “as it sells more, the first question is whether line efficiency and pricing can keep up.”
The cost structure is the least “Buffett-like” part of the business. In 2025, manufacturing overhead accounted for a large share of core business cost. Within the total cost structure for semiconductor wafers, manufacturing overhead accounted for 47.96%, while direct materials accounted for 43.48%, showing that this is a typical heavy-asset, high-depreciation, high-fixed-manufacturing-cost business. By product, in 2025 the direct materials, direct labor, and manufacturing overhead for 300mm semiconductor wafers were about RMB 1.333 billion, RMB 154 million, and RMB 1.317 billion, respectively. Manufacturing overhead for 200mm and smaller semiconductor wafers was also as high as RMB 669 million. This means profits can deteriorate quickly when capacity utilization is insufficient.
If the test is “if the stock market closed for five years, would I want to own this business,” my answer is: I would want to keep tracking the business, but I would not want to hold it heavily at the current valuation as if acquiring a mature enterprise. The reason is not that the business is hard to understand. The reason is that it has not yet proven that the full chain of “capacity expansion - qualification - volume ramp - cost reduction - cash-flow breakeven” has worked.
Industry and Competitive Landscape
Long-term industry demand itself is real. SEMI's 2025 global wafer data showed that global silicon wafer shipment area increased 5.8% year over year to 12,973 million square inches in 2025, but industry revenue fell 1.2% year over year to USD 11.4 billion. SEMI's Q1 2025 disclosure showed that 300mm wafer shipments grew 6% year over year, while 200mm and smaller wafers weakened. This is highly consistent with National Silicon Industry Group's own operating disclosures: its strongest growth came from 300mm, while 200mm and smaller wafers remained under pressure. Long-term demand is not the problem, but pricing, inventory, and structural divergence in the short and medium term will be very clear.
The industry structure is also clear: this is a global oligopoly. In its 2022 annual report summary, the company stated plainly that the world's top five semiconductor wafer manufacturers, Shin-Etsu Chemical, SUMCO, GlobalWafers, Siltronic, and SK Siltron, together held nearly 90% market share. Its 2025 annual report further stated that the top five global manufacturers together held more than 90% market share. In other words, National Silicon Industry Group is not in an industry that is easy to enter or easy to copy. It is in an industry that is hard to enter, but not necessarily easy to profit from even after entry.
In terms of industry stage, this is not a declining industry, nor is it a mature and stable consumer-goods industry. It is a long-term growth, strongly cyclical in the short and medium term, continuously evolving materials industry. 300mm remains the mainstream size. The company has made clear that 300mm offers more than twice the usable area of 200mm and lowers unit chip cost. But 200mm and smaller wafers still have steady demand in power devices, power management, MEMS, automotive electronics, and other fields. Put differently, the industry has structural growth, but not every segment is attractive.
The company's position in mainland China carries weight. The 2024 annual report stated that the company's combined 300mm semiconductor wafer capacity had reached 650,000 wafers per month, that the Shanghai factory had completed 600,000 wafers per month of capacity construction, that annual shipments exceeded 5 million wafers, and that cumulative historical shipments exceeded 15 million wafers. The 2025 annual report further disclosed that combined capacity across two sites had increased to 850,000 wafers per month. This shows that the company has indeed entered the first tier domestically in the 12-inch wafer segment.
But “high industry position” does not equal “strong pricing power.” According to the Q1 2026 disclosure, the company's 300mm wafer sales volume rose sharply, but management explicitly acknowledged that prices declined from the same period of the prior year, and total profit deteriorated further. In the April 2026 investor relations record, management's wording on domestic wafer prices was only that they “will gradually stabilize,” not that clear price increases had already returned. In long-term owner language, this is not a business that has already proven it can pass through inflation and industry volatility smoothly.
So the industry conclusion is neither “an excellent company in a poor industry” nor “a perfect company in a perfect industry.” A more accurate description is: it operates in an upstream materials industry with high strategic value and high technical barriers, but whose commercial returns are not naturally superior; it is a relatively strong domestic company in this industry, but the industry itself does not have consumer-goods-like certainty.
Moat and Management
What the Moat Actually Is
Under a strict Buffett framework, National Silicon Industry Group's moat mainly comes from process qualification barriers, scale and capital barriers, and product portfolio plus technical accumulation, rather than brand, network effects, or channel monopoly. The company's 2025 annual report disclosed that combined 300mm wafer capacity across Shanghai and Taiyuan had reached 850,000 wafers per month. The 2024 annual report also disclosed that after the upgrade projects launched in Shanghai and Taiyuan are completed, the 300mm business will add 600,000 wafers per month on top of the existing base, reaching total capacity of 1.2 million wafers per month, with expected total investment of about RMB 13.2 billion. For later entrants, copying this means copying not just laboratory technology, but also capital expenditure, customer qualification, mass-production management, and delivery systems.
Switching costs exist, but they are not strong enough to block out price competition. Fabs are extremely sensitive to wafer quality, defect density, warpage, thickness uniformity, and surface particle control, and the qualification process itself is a barrier. The company's 2025 annual report disclosed that the Taiyuan factory had completed quality-system audits for multiple customers and started batch sales of prime wafers. 300mm SOI had also entered mass-production implementation and market validation. The process itself shows that changing suppliers is not easy for customers.
But it must be emphasized that the moat is not yet fully reflected in the financial statements. In 2025, revenue from 300mm semiconductor wafers was RMB 2.439 billion, with gross margin of -14.99%. Revenue from 200mm and smaller semiconductor wafers was RMB 1.125 billion, with gross margin of -23.37%. A truly wide and mature moat usually leaves traces in high gross margin, high ROIC, and positive free cash flow. National Silicon Industry Group does not yet have those. It looks more like a moat with barriers, but still on the eve of return realization.
Breaking the moat down item by item makes the picture clearer:
| Item | Judgment | Explanation |
|---|---|---|
| Brand advantage | Weak to moderate | It is not a consumer brand, but in fab qualification systems, “stable mass production, deliverability, and traceability” are forms of credibility. |
| Cost advantage | Insufficient for now | Gross margins for both core products were still negative in 2025, showing that a cost advantage has not been established. |
| Scale advantage | Moderately strong | Domestic 300mm capacity and customer coverage are in leading positions. |
| Network effects | Basically none | This is not a platform business. |
| Switching costs | Moderately strong | Materials qualification and yield ramp mean customer switching is not easy. |
| Channel advantage | Weak | Direct sales dominate; channels are not the core barrier. |
| Patent / process / regulatory barriers | Moderately strong | Technology, clean production lines, qualification cycles, and capital expenditure jointly form barriers. |
| Data advantage | Moderate | Accumulation of mass-production defect control, process parameters, and customer collaboration matters. |
| Corporate culture and operating capability | Moderate | R&D investment continues to rise, but returns are still being validated. |
| Capital allocation capability | Weak | The intensity of capacity expansion has exceeded the speed of return realization. |
This table is a synthesized judgment based on the company's business disclosures, capacity planning, customer qualification, and financial results.
On moat trend, my view is “the technical barrier is widening, but the economic moat has not widened yet.” Technically, the company is indeed expanding 300mm, SOI, epitaxial, and special-specification wafers, while customer qualification is also increasing. Economically, however, pricing pressure and fixed-cost absorption are still squeezing profits. Whether it can cross from “technical barrier” to “financial barrier” depends on three things: 300mm gross margin turning positive, SOI mass production scaling up, and capital-expenditure intensity declining. None of these has been fully delivered yet.
Is Management Trustworthy?
Start with the positives. Management at least has not avoided the problems in public communication. The 2024 annual report directly stated that the two capacity expansion projects mentioned above generated pre-tax losses of about RMB 200 million during the reporting period. The April 2026 investor relations record also clearly explained the three main reasons for larger losses in 2025: depreciation pressure from new capacity, industry price competition and cost pressure, and R&D investment plus asset impairment. This kind of disclosure does not prove excellent capital allocation, but it is at least more candid than companies that only talk about orders and ignore returns.
Now the negatives. This is not a typical “founder-heavy ownership, deeply aligned with shareholders” company. At the end of 2024, the two largest shareholders were the National Integrated Circuit Industry Investment Fund and Shanghai Guosheng Group, with stakes of 20.64% and 19.87%, respectively. By Q1 2026, the major shareholders were still mainly state-owned platforms and industrial funds. The advantage of this structure is resources and industrial coordination. The drawback is that “industrial expansion” can easily come before “per-share returns.” For minority shareholders, this structure naturally lacks the shareholder culture that Buffett tends to prefer most.
Capital allocation is where I have the strongest reservations. In 2025, net operating cash flow was RMB -559 million, but cash paid for purchases and construction of fixed assets, intangible assets, and other long-term assets reached RMB 4.909 billion. Meanwhile, net cash flow from financing activities was RMB 6.472 billion. This shows that the company's current capital-allocation line is not “reinvest existing cash at high returns,” but “use external capital to keep pushing capacity expansion.” This is understandable at the industry stage, but under a value investing framework it means: the company has not yet proven it can use capital to generate returns above its cost of capital.
At the per-share value level, from the end of 2024 to June 2026, shares outstanding increased from 2.747 billion shares to about 3.305 billion shares, up roughly 20.3%. If a company continues to rely on equity financing before free cash flow turns positive, the theoretical “industry opportunity” can be diluted away. Value investors care not about the company becoming larger, but whether intrinsic value per share is growing. On this point, National Silicon Industry Group has not yet delivered a reassuring answer.
So the conclusion on management is: reasonable honesty, clear long-term direction, but capital allocation currently leans more “industry-oriented” than “shareholder-return-oriented.” If cash flow and gross margins improve clearly over the next two to three years, this score can be raised. Until then, I can only give a neutral-to-cautious assessment.
Financial Quality and Owner Earnings
Key Financial Metrics
The table below uses first-hand data from the latest six years to lay out the company's financial trajectory as cleanly as possible.
| Year | Revenue | YoY | Net profit attributable to parent | Net operating cash flow | Gross margin | Weighted ROE | R&D expense ratio |
|---|---|---|---|---|---|---|---|
| 2020 | RMB 1.811 billion | — | RMB 87 million | RMB 377 million | Unknown | 1.14% | 7.23% |
| 2021 | RMB 2.467 billion | 36.2% | RMB 146 million | RMB 307 million | Unknown | 1.47% | 5.10% |
| 2022 | RMB 3.600 billion | 46.0% | RMB 325 million | RMB 459 million | Needs supplemental data | 2.29% | 5.87% |
| 2023 | RMB 3.190 billion | -11.4% | RMB 187 million | RMB -275 million | 16.46% | 1.27% | 6.96% |
| 2024 | RMB 3.388 billion | 6.2% | RMB -971 million | RMB -788 million | -8.98% | -7.07% | 7.88% |
| 2025 | RMB 3.716 billion | 9.7% | RMB -1.508 billion | RMB -559 million | -17.06% | -12.68% | 9.52% |
The data in the table is organized according to company disclosure. 2020-2022 come from the 2022 annual report summary, 2023-2024 from the 2024 annual report, and 2025 from the 2025 annual report. Gross margins for 2023-2025 are calculated from operating revenue and operating cost.
This table says three very important things. First, the company is not without growth; the quality of growth has broken down: 2020-2022 was a phase of “revenue growth, net profit growth, and positive operating cash flow,” while 2023-2025 became a phase where “revenue can still grow, but profit and cash flow deteriorate together.” Second, the collapse in margin is not a small accounting fluctuation, but the result of the industry cycle, pricing pressure, and depreciation from capacity expansion acting together. Third, R&D investment continues to rise, showing that the company is still tackling technology, but this will keep suppressing near-term profit.
Looking only at 2025, financial quality was not ideal. The company generated RMB 3.716 billion of revenue, RMB -1.508 billion of net profit attributable to the parent, RMB -1.774 billion of recurring net profit attributable to the parent, and RMB -559 million of net operating cash flow. In the core business, 300mm semiconductor wafer revenue was RMB 2.439 billion, with gross margin of -14.99%. Revenue from 200mm and smaller semiconductor wafers was RMB 1.125 billion, with gross margin of -23.37%. This means the company's core business had not yet reached positive unit economics in 2025.
Next, leverage and liquidity. At the end of 2025, the company disclosed a current ratio of 1.53, quick ratio of 1.19, debt-to-asset ratio of 40.33%, interest coverage of -6.98, and EBITDA interest coverage of -1.39. At the end of Q1 2026, monetary funds were about RMB 5.789 billion. Short-term borrowings, non-current liabilities due within one year, long-term borrowings, bonds payable, and lease liabilities together were about RMB 9.357 billion, implying net debt of about RMB 3.568 billion. This is not enough to push the company immediately into a liquidity crisis, but for a business still spending heavily on capital expenditure and not yet repaired in profitability, financial flexibility is no longer as comfortable as it was in 2022.
Working capital is also sending pressure signals. At the end of 2025, accounts receivable were about RMB 906 million, slightly lower than at the end of 2024. But inventory rose to about RMB 1.966 billion, clearly above the RMB 1.542 billion at the end of 2024. In an environment of pricing pressure and negative gross margins, rising inventory makes subsequent impairment risk more worth watching. The company's 2025 asset impairment loss reached RMB 694 million, and the annual report explicitly stated that both goodwill impairment and inventory write-down losses were increasing.
On shares and dividends, shareholder returns have not been friendly. The company has not used repurchases as a main tool, and cash dividends are not the main line either. The 2023 dividend proposal totaled about RMB 110 million, but compared with subsequent capital expenditure at a scale of several billion RMB, this is almost negligible. More importantly, share capital has risen noticeably in recent years, showing that the company's growth mainly relies on external capital injection, not internally self-sustaining cash circulation.
Overall, current profit cannot be treated as “high-quality accounting profit,” much less “distributable cash profit.” I have not seen direct evidence of financial fraud in the materials reviewed, and public disclosures do not show debt default. But heavy-asset capacity expansion, rising impairment, negative gross margin, negative operating cash flow, and continuous financing are themselves the financial combination long-term investors should be most alert to.
Owner Earnings Analysis
Under the owner earnings approach, I prefer to start from operating cash flow rather than net profit. In 2025, the company's net operating cash flow was RMB -559 million. In an April 2026 management exchange, management disclosed that group-wide depreciation in 2025 was about RMB 1.3 billion, mainly from the 300mm production line and the Finland 200mm production line. If this RMB 1.3 billion is treated roughly as the minimum maintenance capital expenditure needed to preserve the existing operating capability, then 2025 conservative owner earnings were about RMB -1.86 billion.
Using a stricter free-cash-flow measure, in 2025 the company's “cash paid for purchases and construction of fixed assets, intangible assets, and other long-term assets” was RMB 4.909 billion, corresponding to full-basis free cash flow of about RMB -5.468 billion. This measure clearly includes a large amount of growth capital expenditure, but it also reminds you of a harsh fact: this company is still far from being a real distributable-cash-flow machine.
The 2025 owner earnings can be approximated in the small table below:
| Item | 2025 value | Explanation |
|---|---|---|
| Net profit attributable to parent | RMB -1.508 billion | Already loss-making |
| Net operating cash flow | RMB -559 million | Slightly “less bad” than net profit, but still negative |
| Depreciation for the year | About RMB 1.3 billion | Management figure |
| Estimated maintenance capital expenditure | About RMB 1.3 billion | Conservative assumption: no lower than depreciation |
| Conservative Owner Earnings | About RMB -1.86 billion | Operating cash flow minus maintenance capital expenditure |
| Full-basis free cash flow | About RMB -5.47 billion | Operating cash flow minus total capital expenditure |
The last two rows of the table are estimates based on company-disclosed financial data, not official company metrics. The underlying data comes from the 2025 annual report and management exchange records.
Therefore, the question “what multiple of owner earnings is the current valuation” has no positive answer in a strict sense. If owner earnings are negative, the valuation multiple loses meaning. If one forcibly assumes that the company can someday reach RMB 800 million of normalized owner earnings, then the current RMB 87.38 billion market capitalization is still equivalent to about 109x normalized owner earnings. For an upstream wafer company, that is a very high level of prepayment. This conclusion alone is enough to show that the current price leaves no room for value investors. It is calculated based on the disclosed market capitalization, revenue, and cash-flow data.
Valuation and Margin of Safety
A First Look at Current Valuation
At the Reuters delayed traded price of RMB 26.44 per share on 2026-06-09, National Silicon Industry Group's market capitalization was about RMB 87.38 billion. Using 2025 revenue of RMB 3.716 billion, the current P/S is about 23.5x. Using Q1 2026 net assets attributable to the parent of RMB 16.835 billion, the current P/B is about 5.2x. With profit, free cash flow, and owner earnings all negative, this valuation can only be supported by long-range imagination, not by current cash flow.
Owner Earnings Discount Method
Start with the least flattering method. Because current owner earnings are negative, DCF is extremely sensitive to assumptions. The three scenarios below are already fairly loose scenario assumptions, not intentionally pessimistic suppression:
| Scenario | Revenue CAGR | Owner Earnings margin after ten years | Discount rate | Terminal growth | Implied value per share |
|---|---|---|---|---|---|
| Conservative | 8% | 5% | 12% | 2.0% | About RMB 0.2 |
| Base | 15% | 13% | 10% | 3.0% | About RMB 4.2 |
| Bull | 20% | 20% | 9% | 3.5% | About RMB 13.4 |
The implication of this model is clear: as long as you use “distributable cash flow” rather than “capacity story” to discipline valuation, the current price almost inevitably looks high. The key is not precision to the decimal point, but direction: given current losses and the high dependence of future profit realization on capacity utilization and price recovery, DCF naturally produces a low but honest answer. The model uses 2025 revenue as the base and derives scenarios from current negative owner earnings, different growth rates, and different margin assumptions over the next decade. Base-period data comes from the 2025 annual report and the owner earnings estimate.
Relative Valuation Method
For this kind of company, PE, P/FCF, and EV/EBITDA are not very useful at the current stage because National Silicon Industry Group's earnings and EBITDA are both negative. More useful is the combination of P/S, P/B, and ROE / cash-flow quality. A simple table using peers that can be verified from available data is below:
| Company | Business profile | P/S | P/B | PE | P/CF | ROE / earnings quality | Conclusion |
|---|---|---|---|---|---|---|---|
| National Silicon Industry Group | Domestic 300mm/200mm/SOI | 23.5x | 5.2x | N.M. | N.M. | 2025 ROE -12.68%, deeply negative FCF | Not cheap |
| Leon Micro | Wafers + discrete devices | 10.0x | 5.23x | N.M. | N.M. | More diversified business | National Silicon Industry Group is materially more expensive |
| GlobalWafers | Global pure-play wafer leader | 6.57x | 4.15x | 53.15x | 23.29x | ROE 3.53%, profitable | National Silicon Industry Group's premium is too high |
| SUMCO | Global large-diameter wafer leader | Needs supplemental data | Needs supplemental data | Needs supplemental data | Needs supplemental data | Clear global leader position | Higher quality than National Silicon Industry Group |
The key conclusion from this table is not simply who is cheaper. It is this: National Silicon Industry Group's current price-to-sales ratio is clearly higher than GlobalWafers, a global leader, and also higher than domestic comparable Leon Micro, while its earnings quality and cash-flow quality are worse. This means the logic of “the whole peer group is expensive, so this one is not expensive” does not hold. If relative valuation applies GlobalWafers' 6.57x P/S to National Silicon Industry Group's 2025 revenue, the implied share price is about RMB 7.4 per share. Using Leon Micro's 9.99x P/S gives an implied share price of about RMB 11.2 per share. This already provides a more realistic relative-valuation anchor.
A special reminder: P/B is misleading here. Heavy-asset capacity expansion rapidly increases book assets and net assets. If those assets have not yet generated reasonable returns, a high P/B does not mean asset quality is good, and a low P/B is not necessarily cheap. National Silicon Industry Group's current about 5.2x P/B is close to Leon Micro's 5.23x and above GlobalWafers' 4.15x. With much weaker ROE and free cash flow, this P/B is not only not cheap; it shows the market has already priced in relatively high future execution expectations.
Asset Value and Liquidation Value Method
The asset approach looks somewhat friendlier than DCF, but still does not point to cheapness. In Q1 2026, equity attributable to shareholders of the listed company was RMB 16.835 billion, corresponding to net assets per share of about RMB 5.09. At the end of 2025, book goodwill was about RMB 419 million. Simply deducting goodwill gives tangible net assets per share of about RMB 4.97. This can be viewed as a very rough asset safety cushion.
But this also requires sobriety: these assets include a large amount of highly specialized plants, equipment, and construction in progress. For semiconductor wafer companies, book assets are not cash, and liquidation value is usually below accounting net value. More importantly, the 2025 current ratio had already fallen to 1.53, and net debt had risen. The balance sheet is not so loose that a high valuation can be built entirely on a “liquidation floor.” In other words, book value can explain why the company does not look like a zero, but it cannot explain why it should be worth RMB 26.
Final Intrinsic Value Range and Margin of Safety
Combining the three methods, rather than mechanically averaging them, I arrive at the following ranges:
| Range | Value per share | Basis |
|---|---|---|
| Conservative intrinsic value range | RMB 4-6 | Mainly based on tangible net assets, liquidation discount, and weak cash-flow reality |
| Reasonable intrinsic value range | RMB 7-11 | Mainly based on peer P/S anchors and a domestic-substitution scenario with moderate success probability |
| Bullish intrinsic value range | RMB 12-16 | Requires successful scale-up of 300mm / 300mm SOI, sustained gross-margin repair, and lower financing pressure |
At the current RMB 26.44, the market price is roughly 341%-561% above the conservative range and 140%-278% above the reasonable range. Even against my bullish range, it is still 65%-120% higher. This means there is currently no margin of safety.
Therefore, my price bands are:
| Price range | Judgment |
|---|---|
| RMB 5-8 | Ideal buy price range |
| RMB 8-12 | Acceptable long-term holding price range |
| RMB 12-18 | Understandable as a trade, still expensive on value |
| Above RMB 18 | Clearly overvalued range |
The most fragile assumption in the valuation is only one: you must believe that 300mm expansion will ultimately produce sufficiently high utilization, sufficiently good yield, and a sufficiently stable pricing environment, allowing gross margins, operating cash flow, and owner earnings to turn positive and keep improving together. If volume rises but pricing and cost do not improve in tandem, then so-called “domestic substitution” may create revenue for shareholders, but not value.
Risks, Comparisons, and Checklist
Risks and the Strongest Bear Case
The most important risk is not share-price volatility, but permanent capital loss. The first category is competition and pricing risk: the five global wafer giants still control the main market, while domestic players are also accelerating capacity expansion. In Q1 2026, the company already disclosed that 300mm sales volume rose sharply but prices still fell from the same period of the prior year. The second category is technology execution risk: if qualification of SOI, special specifications, high-end epitaxial wafers, and heavily doped products progresses more slowly than expected, depreciation pressure from capacity expansion will continue to consume profits. The third category is capital allocation risk: if operating cash flow and gross margins repair more slowly than expected, the company may keep relying on debt and equity financing, eventually turning “industry opportunity” into “per-share value dilution.”
The fourth category is cycle risk and drag from the 200mm business. SEMI and the company point to the same reality: 300mm is recovering, but 200mm and smaller wafers are recovering more slowly. In the April 2026 management exchange, management also stated plainly that 8-inch epitaxy was close to full utilization, but polished products were still affected by weak mobile-phone and automotive industries and faced substantial market pressure. As long as the 200mm business does not stop declining, group-level margin repair will be slowed.
The fifth category is accounting and impairment risk. I have not seen direct evidence of monetized fraud in the materials collected, but 2025 asset impairment losses already reached RMB 694 million, and the annual report stated that goodwill impairment and inventory write-down losses were both increasing. For a capacity-expanding heavy-asset company, impairment is not a small issue. It often means the economic returns on some past capital investments were below the original expectations.
The strongest bear case is actually simple: this may be a company with the right technology direction and important strategic position, but whose shareholder returns remain below the cost of capital for a long time. Investors bearish on it do not necessarily deny domestic substitution, nor the growth of 300mm. They are more likely focused on three other facts: negative gross margin, negative free cash flow, and excessive valuation. If in the next few years the company keeps expanding capacity, keeps winning orders, and also keeps losing money, the market will eventually discover that “scale” does not equal “value.” This is the largest permanent capital loss scenario, based on the company's financials over the past two years and the current valuation.
Comparison with Other Opportunities
Compared with the strongest global competitors, National Silicon Industry Group's industrial position is respectable, but its economic quality gap is clear. SUMCO is a world-leading large-diameter wafer manufacturer, and GlobalWafers is also a global leading pure-play wafer company. Yet National Silicon Industry Group's current P/S is higher than GlobalWafers', while ROE and cash flow are worse. As a long-term owner, if you can allocate capital freely across the global investable universe, National Silicon Industry Group currently does not show why it deserves a higher price than mature global leaders.
Compared with a broad-based index, National Silicon Industry Group's problem is not the absence of upside imagination. The issue is that the investment case depends on multiple difficult conditions materializing together: 300mm gross margin turning positive, SOI scaling, 200mm recovering, capital expenditure declining, and financing slowing. A broad-based index may not make you rich overnight, but it does not force you to bet portfolio returns on one production line, one material, one cycle, and one refinancing. For balanced, conservative investors, I do not think buying National Silicon Industry Group at the current price is clearly better than holding a broad-based index fund. This is based on the company's operations and valuation discussed above.
Compared with high-grade bonds or risk-free assets, the issue is even more direct: when owner earnings are still negative, you do not receive a certain cash return and can only hope for a future inflection point. In other words, buying this stock today gives you a “long-dated option,” not a “current cash machine.” If your risk preference is balanced and conservative, this payoff structure is not ideal. This is based on the company's current owner earnings and free cash flow status.
If I could hold only five assets, my answer is: it does not qualify for the portfolio. Not because the industry is unimportant, but because at the current price it consumes your capital without offering a certainty of return that matches its risk. For long-term value investors, capital should be allocated first to companies or index assets that are understandable, have more stable cash flow, and are valued with more restraint. This is based on the full analysis above.
Investment Checklist
| Checklist item | Conclusion | Explanation |
|---|---|---|
| Can I understand this business? | Pass | A qualification-based wafer materials business selling to fabs; the logic is clear. |
| Does it have long-term stable demand? | Pass | Long-term demand exists, but cyclicality is obvious. |
| Does it have a durable moat? | Uncertain | It has technology and qualification barriers, but the financial moat has not been realized. |
| Does it have pricing power? | Fail | Q1 2026 volume growth with pricing pressure already illustrates the problem. |
| Can it generate stable free cash flow? | Fail | FCF was deeply negative in 2024-2025. |
| Are its returns on capital excellent? | Fail | ROE has turned negative in recent years. |
| Is management trustworthy? | Basically pass | Communication is relatively candid, but that does not equal excellent capital allocation. |
| Is capital allocation rational? | Uncertain, leaning fail | The capacity expansion direction is reasonable, but shareholder returns have not been validated. |
| Is the balance sheet robust? | Uncertain | It can still support operations, but the safety cushion is thinning. |
| Is valuation below intrinsic value? | Fail | The current price is materially above the cross-checked range from three methods. |
| Is the margin of safety sufficient? | Fail | Basically none. |
| Would I feel comfortable holding it long term? | Fail | Cash-flow and refinancing risks remain high. |
| What facts would make me sell? | Clear | Gross margin failing to turn positive for a long time, continued large dilution, qualification failure, or cash-flow deterioration. |
| Am I tempted to buy only because the share price has risen or sentiment is strong? | Requires self-discipline | Current valuation clearly contains long-range expectations. |
The checklist judgments above are based on a synthesis of company annual reports, quarterly reports, management exchanges, and current market valuation.
Final Investment Conclusion
Final Rating
Watch
One-Sentence Investment Thesis
National Silicon Industry Group is a domestic 300mm wafer leader worth tracking over the long term, but before it forms stable positive free cash flow and positive owner earnings, the current price is buying expectations more than value.
Core Bull Points
The company has established a clear position in the domestic 300mm semiconductor wafer segment. In 2025, combined capacity across two sites reached 850,000 wafers per month, and the company continues to advance toward the 1.2 million wafers per month target.
The customer base is high quality, covering major domestic fabs and already serving multiple international customers. Once qualification and supply relationships stabilize, stickiness is not low.
In Q1 2026, revenue increased 35.22% year over year, and 300mm wafer sales volume increased by more than 90% year over year, showing that demand and the domestic-substitution trend are indeed advancing.
High-end product lines such as 300mm SOI, heavily doped wafers, and epitaxial wafers are moving through mass production and validation. These are potential sources of future gross-margin improvement.
Core Bear Points
Gross margins for both core wafer products were negative in 2025, showing that scale expansion has not yet become economic return.
Operating cash flow remained negative in 2024-2025, and 2025 capital expenditure was still as high as RMB 4.909 billion, leaving free cash flow deeply negative.
In 2025, the current ratio fell to 1.53, the debt-to-asset ratio rose to 40.33%, and interest coverage was -6.98, indicating rising financing dependence.
The current valuation is about 23.5x P/S and 5.2x P/B, clearly above what current earnings and cash-flow quality can support.
The share count has risen noticeably in recent years, indicating that shareholders may continue to bear dilution costs.
Key Assumptions
For this investment to work in the future, at least the following conditions must be met: 300mm gross margin continues to turn positive; SOI and high-specification products materially increase their revenue contribution; the 200mm-and-below business stops declining; operating cash flow turns from negative to positive in 2027-2028; and future financing intensity declines clearly, avoiding continued dilution of per-share value. None of these assumptions has yet been proven by facts. This judgment is based on the company's currently disclosed operating data and management exchanges.
Reasonable Buy Price
RMB 5-8 per share. The basis mainly comes from three points: first, the asset floor of Q1 2026 net assets per share of about RMB 5.09 and tangible net assets per share of about RMB 4.97; second, relative valuation anchors based on GlobalWafers and Leon Micro P/S imply a roughly reasonable price for National Silicon Industry Group of RMB 7-11 per share; third, even under a more optimistic owner earnings scenario, the upper end of valuation still struggles to support the current price.
Target Holding Period
If the future price returns to a range with a sufficient margin of safety, and operating data begins to confirm improvements in gross margin and cash flow, the holding period should be viewed as 5-10 years. But at the current price, I would rather keep it on the watchlist than buy it directly as a ten-year compounding asset. This is based on the company's industry stage and current valuation.
Expected Annualized Return
The following returns are rough ranges assuming purchase at the current roughly RMB 26.44, a ten-year holding period, and eventual market repricing according to the corresponding operating scenario, without assuming meaningful dividend contribution:
| Scenario | Expected annualized return |
|---|---|
| Conservative | -12% to -15% |
| Base | -5% to -8% |
| Bull | 0% to +5% |
The reason even the bull-case return is not high is not that the company lacks room to grow, but that the current price already reflects too much long-range growth in advance. This conclusion is derived from the intrinsic value range above compared with the current price.
Maximum Loss Risk
In my view, in a worst-case scenario, 60%-80% permanent capital loss is not impossible. In an extreme case, if 300mm capacity utilization stays below breakeven for a long time, 200mm continues to drag, impairment rises, refinancing continues to dilute, and the market ultimately prices the company using asset value or a low P/S multiple, a share price returning to the RMB 5-10 range would not be absurd. The basis comes from current book net assets, negative free cash flow, and the high valuation.
Tracking Indicators
The most important items to track in the future are not share prices, but the following operating indicators:
300mm wafer sales volume, utilization, and yield.
When 300mm and overall gross margins turn positive.
Net operating cash flow and full-basis capital expenditure.
The number of mass-production customers for 300mm SOI, its revenue contribution, and gross-margin performance.
Price, orders, and inventory improvement in the 200mm-and-below business.
Inventory size and asset impairment losses.
Current ratio, debt-to-asset ratio, and interest coverage.
Share capital changes and refinancing arrangements.
Signals That Would Trigger Reassessment
If any of the following occurs, I would immediately reassess the investment logic:
300mm wafer gross margin remains negative for several consecutive quarters with no sign of improvement.
Operating cash flow still cannot turn sustainably positive after 2027.
Mass-production progress for SOI, high-specification products, and heavily doped products is clearly slower than management's description.
The company carries out another large equity financing or material dilution.
Inventory and impairment keep rising, showing that the mismatch between capacity expansion and demand has not eased.
Major customer qualification fails, or overseas expansion is blocked.
Interest rates, exchange rates, or the debt structure deteriorate, causing financial expenses to rise materially.
All of these signals map directly to this report's core assumptions.
Final Recommendation
Calmly put, National Silicon Industry Group is not uninvestable; it is just not worth investing in today through a “long-term value investing” lens. If you are an industrial-trend investor, you can keep studying it and wait for substantial inflection points in gross margin, cash flow, SOI mass production, and financing intensity. If you are a balanced, conservative investor targeting ten-year compounding, my recommendation is: do not buy yet; wait patiently for a lower price or clearer operating execution. At today's price, you are taking on high-difficulty execution risk without receiving enough margin of safety.
This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.
Full report
Sign in to read the full report
Sign up free to unlock the full text, the Baillie growth scorecard, and full-text search.
Log in / Sign up free