Hua Hong Semiconductor: Value and Price in Mature Specialty Process Platforms

That origin shaped the later path toward increasing focus on mature and specialty processes, rather than "doing everything." The Hua Hong NEC name remains separately defined as a predecessor entity in the prospectus, showing that the lineage was never erased. Hua Hong's later specialty was built around dedicated platforms for power, eNVM, analog, power management, logic, and RF, not an advanced-node race. That path reflected resource constraints: compared with TSMC, Samsung, and Intel, Hua Hong lacked both the most advanced equipment conditions and the large-customer base for the most advanced nodes. It also reflected the reality of China's local demand structure: many industrial, automotive, power, MCU, and security chips do not need the most advanced process. They care more about reliability, cost, stable supply, and localized manufacturing.

2014 was Hua Hong's first true move into international capital markets. The prospectus shows that in October 2014 the company issued 228,696,000 shares at HKD 11.25 per share, listed on the main board of the Hong Kong Stock Exchange on October 15, and raised about USD 320.2 million in net proceeds. The story it told the market was clear: China-based, specialty-process, pure-play foundry. At the time, that story was not glamorous, because global capital markets preferred advanced processes, the smartphone chain, and high-growth design companies. Hua Hong was therefore long stuck in a slightly awkward position: a manufacturer that truly made money and truly expanded capacity, but not a company at the center of the global narrative.

The National Integrated Circuit Industry Investment Fund's investment in 2018 was the second important turning point. The prospectus disclosed that the fund subscribed for 242,398,925 shares in 2018 at HKD 12.9002 per share. The meaning of that capital was to push Hua Hong toward 12-inch lines and a more complete specialty-process platform, rather than merely to fill a funding gap. After that, Hua Hong gradually moved from an "old-line specialty foundry mainly based on 8-inch" to a "platform company with both 8-inch and 12-inch capacity." Fab 9 is the culmination of that stage. Management later disclosed at the 2026 earnings briefing that total capital investment in the Fab 9 project is about USD 6.7 billion; by the end of 2025 more than USD 5.0 billion had already been spent, and most of the remaining USD 1.2-1.3 billion cash outflow would occur in 2026, with a smaller portion extending into 2027. That number completely rewrote Hua Hong's business story: it is now using very large capital expenditure to buy a future capacity step-up, rather than operating as a conservative mature factory.

The 2023 return to the A-share market was the third narrative upgrade. Hua Hong listed on the STAR Market at CNY 52 per share, issued about 408 million shares, and raised about CNY 21.203 billion, making it one of the largest A-share IPOs of that year. Hong Kong valued it in the traditional framework of "specialty-process manufacturing." A shares gave it part of the premium for "domestic substitution + high-end manufacturing + scarce foundry asset." This dual-pricing structure later became the key to understanding Hua Hong's valuation: the same company faces different risk appetites, capital structures, and policy imagination in H shares and A shares. By 2026-06-12, the H-share price translated into USD was about USD 17.77, while the A-share price translated into USD was about USD 32.47, implying an A-share premium of about 83% over the H shares.

From 2025 to 2026, Hua Hong entered the most complex phase of the current cycle. Full-year 2025 revenue was USD 2.4021 billion and average capacity utilization was 106.1%, showing that demand was genuinely strong. Fab 9 ramped quickly, and the Fab 5 acquisition will bring related 55/40nm capacity and process platforms into the listed company. But the problems arrived at the same time: low-teens gross margin, persistently high capital expenditure, and the future need to advance Fab 9's second phase. Over the past two decades, Hua Hong has proved that it can build a line, deepen specialty processes, and seize windows between policy support and industry cycles. Today's problem is a different version: how to turn "larger scale" into genuinely better margins and cash flow during China's wave of mature-node expansion.

Longitudinal Financial Review

Hua Hong's financial path over the past several years looks like a curve that first surged, then fell hard, then climbed slowly. At the semiconductor cycle peak in 2021, revenue reached USD 1.631 billion, up 69.6% year on year; gross margin was 27.1%; profit attributable to owners of the parent was USD 320 million. In 2022, profit attributable to owners of the parent rose further to USD 449.9 million. During that period, tight industry supply, firm mature-node pricing, and real customer shortages allowed Hua Hong, as scarce local foundry capacity, to capture one of the best profit windfalls in its history.

The turning point came after 2023. Profit attributable to owners of the parent fell to USD 280.0 million in 2023, a clear decline from 2022. By 2025, revenue had grown again to USD 2.4021 billion, but gross margin was only 11.8%. This shows that Hua Hong's financial changes over the past two years can no longer be explained simply by "weak cycle." The real pressure came from two places: first, mature-node pricing and customer mix were normalizing; second, the ramp of new 12-inch lines lifted depreciation and fixed costs, so revenue recovery did not mean profit recovery at the same pace. Put together, these explain why the core feature of 2025 was "clear revenue recovery, slow profit recovery."

Cash flow is more informative than the income statement. The A-share annual report shows that 2025 net cash flow from operating activities was CNY 5.065 billion, up 40.38% year on year. In the same year, cash paid for the purchase and construction of fixed assets, intangible assets, and other long-term assets was CNY 12.931 billion, far above operating cash flow. Translated at the 2025 year-end USD rate, operating cash flow was about USD 721 million, capex cash flow was about USD 1.840 billion, and rough free cash flow was about negative USD 1.12 billion. In other words, Hua Hong was not short of orders or shipments in 2025, but it remained a company putting large amounts of cash back into fabs.

This is why valuing Hua Hong on net profit can be highly misleading. 2025 net profit attributable to the parent was only CNY 377 million, very far from CNY 5.065 billion of operating cash flow. Part of the gap comes from heavy non-cash costs such as depreciation, which compress accounting profit. Another part is that even when operating cash flow is decent, the true "owner earnings" left to shareholders remain limited, because factories are still expanding, equipment is still being installed, and Fab 9 still needs cash. If 30%-40% of 2025 capex is treated as maintenance capex, Hua Hong's owner earnings are roughly between negative USD 2 million and USD 169 million, implying a yield of about -0.04% to 0.46% on the current total market capitalization. This explains the problem better than the headline H-share static P/E above 500x: the company's cash-return rights are temporarily locked in future fab construction, rather than the company having no value.

The balance sheet can still hold for now. At the end of 2025, cash and cash equivalents were about CNY 34.398 billion. In Q1 2026, total assets were about CNY 101.949 billion, long-term borrowings were about CNY 21.349 billion, and non-current liabilities due within one year were about CNY 5.649 billion. Comparing cash with the sum of these two interest-bearing liabilities, there is no immediate balance-sheet liquidity crisis. Hua Hong's problem is more about return on capital than debt service. For a company like this, the truly dangerous situation is "continuously investing new money while returns on new capacity are below the cost of capital," rather than "unable to borrow."

One accounting-basis issue must be called out separately. The product breakdown for principal operations in the A-share annual report gives a 2025 "gross margin of principal operations" of 18.72%, while the chairman's letter to shareholders gives full-year gross margin of 11.8%. This looks more like a statistical-scope difference than a simple typo: the former is closer to the gross margin at the principal-operation level under Chinese reporting rules, while the latter is the consolidated gross margin the company consistently discloses externally in quarterly reporting. In the valuation and quarterly tracking below, I put more weight on the latter because it is continuous with the Q4 13.0% and Q1 13.0% historical sequence. For cash flow and the balance sheet, I use the A-share annual report as a supplement.

Share-Price and Valuation History

Hua Hong's capital-market history can be divided into three stages. The first was the long quiet period after the 2014 Hong Kong listing. The company listed in 2014 at HKD 11.25 and raised about USD 320.2 million in net proceeds. For a long time, Hua Hong was not a semiconductor stock that global capital watched closely, because it sat between advanced processes and high-growth design companies.

The second stage was 2021-2023: "profit delivery first, valuation pullback later." In 2021, the cycle peaked and both revenue and gross margin surged. Profit continued to rise in 2022. In 2023, profit began to fall, and the market relearned that mature-node foundry is not a business of permanent shortages. The 2023 A-share listing lifted the valuation center once, but that uplift looked more like liquidity and policy-label repricing than a return of margins to peak levels.

The third stage is the narrative reconstruction from 2025 to 2026. Q4 revenue reached USD 659.9 million, a record high, and Q1 2026 revenue reached USD 660.9 million. In May 2026, the Hong Kong shares briefly climbed to a 12-year high of HKD 135.6, after which the 52-week range widened to HKD 29.4-177.0. The most common misread in this round is to treat "share-price rise" as "profit already delivered." In reality, the market is paying in advance for Fab 9, Fab 5, localized manufacturing, and specialty-process pricing.

Using today's data, Hua Hong has been lifted by the market from the framework of a "traditional cyclical manufacturer" into the valuation framework of a "scarce platform asset." H-share market pages show that the company currently pays no dividend, and trailing P/E on the H-share basis exceeds 500x. Based on H-share 2025 year-end net asset value per share of USD 3.806, the current H-share price corresponds to about 4.7x P/B. On an A+H combined market-capitalization basis, the company's total P/S is about 15.3x and total P/B is about 5.6x. This means Hua Hong's current pricing logic has moved away from the traditional mature-foundry method of "looking at profit" and now looks more like asking whether process platforms and localization share can be re-rated over the next three years.

Business Model and Moat

Hua Hong's revenue structure is concentrated, but it is concentrated in several repeatedly validated technology platforms rather than in a single customer. The company website lists the core platforms as embedded/standalone non-volatile memory, power devices, analog and power management, logic, and RF. By Q4 2025, embedded NVM revenue was USD 180.2 million, or 27.3%; analog and power management was USD 173.8 million, or 26.3%; and power devices were USD 168.9 million, or 25.6%. These product lines together account for most revenue. Hua Hong's real commercial machine is to let large numbers of mature but sticky industrial, automotive, power, and control chips iterate over the long term on its process platforms, rather than betting on one hit chip.

The benefit of this model is diversified demand and long platform life. The downside is also clear: pricing power cannot be as strong as in advanced nodes, and gross margin is more easily swallowed by new depreciation and changes in industry supply and demand. Management itself put the issue plainly at the Q1 2026 results briefing: fab utilization ultimately depends on technology competitiveness and price; if pricing is too high, utilization will fall. For a company like Hua Hong, platform depth and pricing discipline must hold at the same time. If either side slips, financial leverage will immediately hit the income statement.

Its real moat mainly has three parts.

First is the accumulated depth of specialty-process platforms. Hua Hong has worked for nearly thirty years in eNVM, power, analog power, and logic/RF. Management said explicitly at the earnings briefing that Hua Hong may be first domestically in several platforms and is also very competitive internationally. This kind of moat is less visible than a consumer brand. It shows up in PDKs, IP, yield, automotive-grade validation, customers' second tape-out experience, and a full set of engineering collaboration capabilities. A customer making MCUs, power devices, or PMICs will not easily move fabs just because another provider is 5% cheaper.

Second is switching cost. Many of Hua Hong's customers are system companies and fabless chip designers that need stable long-term supply, rather than the most advanced and expensive chip companies. For these customers, changing foundries requires recalculating the validation cycle, BOM, reliability testing, automotive certification, and downstream customer requalification, not just wafer price. Hua Hong's ability to deepen platforms is therefore more important than making nodes more advanced. The Q4 2025 product-mix change also proves this point: embedded NVM and analog and power management grew significantly faster than power devices, showing that the company is migrating toward stickier specialty platforms, instead of simply relying on the power categories most exposed to price competition.

Third is its localization position in a geopolitical environment. Management specifically mentioned that European customers, and now some U.S. customers, are advancing "China for China," moving certain products that were originally manufactured overseas back to China to find local manufacturing partners. When these customers look for China-based partners, they usually put more weight on technology base, scale, and stability, so Hua Hong often becomes one of the first choices. This is not a permanent license, but it will be a very real order source for Hua Hong over the next few years.

The moats that do not exist should also be stated clearly. Hua Hong has no advanced-node barrier, no TSMC-like ecosystem pressure, and has not yet become a mature-node high-cash-return machine like UMC. The place where the market most easily overstates the company is turning "local scarcity" into "permanent high profitability." If China's 12-inch mature-node supply keeps coming online, Hua Hong's specialty-platform depth will still have value, but valuation and margin will be hard to keep high forever at the same time.

On governance, Hua Hong's base color is typical state-asset control. The A-share filing materials explicitly identify the company as a listed company under Hua Hong Group and under de facto state control, with Hua Hong International and Hua Hong Group at the core of the control chain. This structure has two sides: financing and policy resources are usually steadier, and support is stronger when large projects advance; the cost is that the governance discount is hard to erase completely, especially around related-party transactions. The acquisition of Huali Micro, where Fab 5 sits, is a typical related-party restructuring. The counterparties include Hua Hong Group, Shanghai Integrated Circuit Industry Investment Fund, the second phase of the National Integrated Circuit Industry Investment Fund, and other related parties or industrial capital. For ordinary shareholders, such transactions require ongoing monitoring of synergy delivery and whether the consideration is fair, rather than being impossible in principle.

Management has also changed materially in recent years. In the 2023 annual report, the chairman was Zhang Suxin and the president was Tang Junjun. By Q1 2026, the quarterly report was signed by Bai Peng as chairman of the board and president. This change is not necessarily negative by itself, but it is enough to remind investors that Hua Hong is not a founder-dominated company. It is closer to a large manufacturing platform jointly driven by state-owned shareholders, professional managers, and industrial policy. When tracking this company, management credibility should be judged by delivery against quarterly guidance and capital-allocation discipline, not by narrative appeal.

Industry and Cycle Analysis

The first important fact when placing Hua Hong back into the industry is that foundry remains a highly concentrated, heavy-asset industry with very high capital barriers. Public credit research citing TrendForce data says the top ten global foundries took 97% of market share in Q3 2025, including TSMC at 71.0%, SMIC at 5.1%, UMC at 4.2%, GlobalFoundries at 3.6%, Hua Hong at 2.6%, and VIS and Nexchip each at 0.9%. This means Hua Hong is not a global giant, but it is also not a marginal player. It is already in the position of "inside the global top ten, but still clearly below the top five."

The second fact is that industry growth is not evenly distributed. WSTS's spring 2025 view confirmed that semiconductor market growth would continue into 2026, with the global semiconductor market reaching USD 346.0 billion in the first half of 2025, up 18.9% year on year. But the center of growth is still more in AI, memory, and advanced logic. The parts of mature nodes that truly benefit are the "pulled by AI but not directly AI main-chip" areas: AI peripherals, power, connectivity, automotive-grade chips, and industrial control. Hua Hong management also admitted in Q&A that the most direct AI-related demand pull for the company is power management, followed by MCU and power devices, and that the impact is not as concentrated as in advanced GPUs.

The third fact is that supply-side expansion is pushing the industry back into a test of discipline. SEMI expects global 300mm fab equipment spending to rise to USD 133.0 billion in 2026 and USD 151.0 billion in 2027. Another SEMI dataset also says China is expected to continue leading with USD 94.0 billion of 300mm equipment investment in 2026-2028. This is good and dangerous for Hua Hong at the same time. It is good because the company itself is part of China's 12-inch expansion, and local policy support plus equipment and materials localization will give it a longer runway. It is dangerous because China's mature-node expansion is not being done by Hua Hong alone. Once the industry overbuilds, the whole sector comes under pressure together.

Hua Hong therefore belongs to several cycles at once: the semiconductor cycle, inventory cycle, capital-expenditure cycle, and policy cycle. In an upcycle, the most beneficial variables are always utilization and ASP. In a downcycle, the most fragile variable is gross margin, because fab depreciation, labor, energy, and maintenance are hard to reduce in line with revenue. The larger Fab 9's capex, the larger this operating leverage becomes. Hua Hong's 2025 operating cash flow rose while net profit still looked weak. Fundamentally, that is the simultaneous existence of an upcycle and rising depreciation.

Policy and geopolitics affect Hua Hong in a way that does not immediately get stuck on EUV as advanced nodes do, but they are not background noise. On one hand, the company itself said at the 2026 earnings briefing that future new projects will purchase a higher proportion of domestic equipment, showing that management already treats equipment and materials localization as a long-term direction. On the other hand, overseas customers' "China for China" and Taiwanese peers' observation that "U.S. customers are moving orders back to Taiwan" both show that geopolitics is dividing mature-node orders into regional supply chains. Hua Hong's opportunity is to become a first-tier carrier of China-based mature nodes and specialty processes. Its risk is that if regionalized supply chains solidify, the global valuation framework will lean more toward manufacturers with multi-region footprints across the U.S., Europe, and Taiwan.

Horizontal Competitor Analysis

If Hua Hong is placed among global mature and specialty-process foundries, it faces three very different groups of competitors rather than one opponent. The first is SMIC, larger in scale, broader in nodes, and even with some advanced-node options, making it the ceiling reference for mainland China foundry. The second is mature-node incumbents such as UMC, VIS, and GlobalFoundries, whose common traits are more mature margins, cash flow, and capital discipline than Hua Hong. The third is mainland China 12-inch challengers such as Nexchip, which may not have more advanced nodes but are expanding quickly and have younger cost curves.

Table note: Hua Hong valuation is calculated using the 2026-06-12 A/H closing prices, issued share counts, 2025 revenue, and net asset value per share. Other valuation and profit-quality metrics come from company websites, annual reports, or market snapshots. The indicators are not all on the same basis; the purpose is to see relative position rather than run a mechanical same-basis regression.

The most jarring point in the table is that Hua Hong has the lowest margin but the valuation is not low. On a combined basis, it is even much more expensive than mature-foundry peers UMC, VIS, and GF. Hua Hong's current high valuation reflects the market treating it as a "scarcer China-based specialty-process platform," not as a "more profitable mature foundry." This also explains why Hua Hong is already not cheap when looking only at the H shares; once the A shares are aggregated, the story premium becomes more obvious.

SMIC is Hua Hong's "ceiling reference." Its advantage is not only scale, but a complete technology spectrum, and the market is willing to grant higher tolerance for broader node coverage and policy scarcity. Customers often choose SMIC because it can cover a fuller set of process routes and holds an irreplaceable strategic position in mainland China. The real difference between Hua Hong and SMIC is "who offers broader manufacturing spectrum capability," not "who is more Chinese." Hua Hong is more focused in specialty processes, while SMIC is stronger in technology bandwidth and customer mindshare.

UMC has become another kind of company. UMC's 2025 revenue was TWD 237.6 billion and gross margin was 29.0%. The market prices it much more like a mature-node "cash machine" than an "expansion story stock." Customers often choose UMC for mature-node process stability, yield, long-term platform supply, and capital discipline. It may not offer the flashiest node, but it is more like a highly predictable, low-drama manufacturing partner. Compared with UMC, Hua Hong's biggest shortcoming becomes clear: the platform narrative already looks similar, but financial quality does not.

VIS and GlobalFoundries provide two other comparisons. VIS is smaller, but 2025 revenue was TWD 48.591 billion, ROE was 12.4%, and 4Q25 gross margin was 27.5%, making it look much more like the profit profile a mature-node specialty player should have. VIS itself is also warning investors that China's mature nodes are building "excess capacity" and may trigger vicious price competition. GF is the globalized version of mature and specialty processes: 2025 revenue was USD 6.791 billion and gross margin was 24.9%, with automotive, defense, and communications orders supporting its multi-region capacity logic. Customers choosing GF are buying geographic diversification and supply-chain security. Customers choosing Hua Hong are buying China-based implementation and specialty platforms. The certainty being sold is not the same.

Nexchip looks like Hua Hong's nearest "underfoot rival." It has mass-produced platforms from 150nm to 40nm; its 28nm OLED products remain under validation, and its 28nm logic platform has completed development. Hong Kong prospectus materials also state that it was the world's ninth-largest and mainland China's third-largest foundry in 2025. This company represents the rapid emergence of a second tier in China's local 12-inch mature foundry, not advanced-node substitution. Customers choose Nexchip more for 12-inch mature platforms, display drivers, power management, and cost efficiency. If Hua Hong cannot keep deepening its specialty-process barriers, the first competition it feels in the future may not come from TSMC, but from these younger local fabs with newer equipment and longer depreciation curves.

Hua Hong's position in this ecosystem is therefore closer to a "first-tier platform company among local specialty-process challengers." It is not a global leader, but it is not a simple follower either. The profit pools it most directly competes for are China-based power, MCU, eNVM, and PMIC. The players most likely to take profit pools back from it are SMIC's broader platform capability and local 12-inch capacity such as Nexchip. If the industry sees technology substitution, price war, or demand retreat, whether Hua Hong's position weakens will depend on whether "specialty process" shows up in margins, not only in presentations.

Current Fundamentals and Bull-Bear Divide

Based on the quarterly sequence that can be directly checked from primary materials, the key phrase for Hua Hong's current fundamentals is "revenue stepping up, margin repairing slowly." Q1 2025 revenue was USD 540.9 million and gross margin was 9.2%; at the time, the company's Q2 guidance was only USD 550-570 million of revenue and 7%-9% gross margin. By Q3 2025, revenue had risen to USD 563.9 million and gross margin to 13.2%. Q4 revenue then reached USD 659.9 million and gross margin was 13.0%. Q1 2026 revenue was USD 660.9 million, up 22.2% year on year, and gross margin remained 13.0%, though it had improved by 3.8 percentage points year on year. This cadence shows that orders and shipments have clearly improved, while margin recovery is still stuck in the low teens.

This recovery is being driven by several platforms together, not by a single product. In the company's Q4 2025 revenue structure, embedded NVM grew 31.3% year on year, analog and power management grew 40.7%, logic and RF grew 19.2%, while power devices grew only 2.4%. This matters because it shows Hua Hong is lifting revenue through higher-value-added platforms with deeper customer binding, rather than relying on the most price-war-prone category of "pure power devices." Management further noted at the Q1 2026 earnings briefing that AI's most direct pull for the company comes mainly from power management, with MCU and power devices also contributing, but more as peripheral beneficiaries.

The themes now being traded by the market can be split into four layers. The first is cyclical recovery: high utilization and recovering end demand. The second is localization: European customers and some U.S. customers are advancing "China for China," creating real demand for orders to move into China-based manufacturing. The third is capacity synergy: Fab 9 reaches peak capacity this year, while Fab 5 adds 55/40nm specialty-process capacity. The fourth is valuation recovery: capital has begun to view Hua Hong as one of the few foundry platforms in A shares and H shares that can absorb China's mature-node domestic substitution. Real fundamentals support the first two layers, the third is being delivered, and the fourth is the most prone to overheating.

The core bullish evidence has three parts. First, average capacity utilization reached 106.1% in 2025, at least showing that this is not "building fabs with no one to use them." Second, Fab 5 will bring 55/40nm specialty processes directly into the listed company, solve part of the horizontal-competition issue, and expand customer acceptance capacity. Third, localized manufacturing demand is real, and Hua Hong already has accumulated capability in domestic specialty-process platforms; orders are not just short-term sentiment.

The core bearish evidence is equally solid. First, gross margin is still only around 13%, far from the truly comfortable profit range in mature foundry. Second, although 2025 operating cash flow was strong, free cash flow remained deeply negative, and Fab 9 cash outflows will continue in 2026. Third, China's 300mm equipment investment is still advancing at high intensity, and the next round of pressure may come from supply, not necessarily demand. Bulls are betting on platform value and localization share. Bears are watching margins and cash flow. The current market price looks more like it is pricing the former.

Valuation Analysis

Conclusion first: Hua Hong is now least suitable for a single P/E lens, and best analyzed through "owner earnings after cash-flow look-through + P/B + relative P/S" together. The reason is simple. The income statement is compressed heavily by depreciation, and capital expenditure is large enough to make free cash flow materially negative. H-share market pages show trailing P/E above 500x and no dividend. If investors mechanically use that number, they will reach the surface conclusion that it is "absurdly expensive," while missing the expansion-period distortion behind accounting profit and cash flow. But if free cash flow and owner earnings are added back into the analysis, the story still does not look very cheap.

Historical and Peer Valuation

The most important references for Hua Hong's current valuation are mature and specialty foundry peers, not the company's own past traditional cyclical-stock center. UMC's current P/S is about 7.03x, GF's about 6.59x, and VIS's about 6.53x. SMIC's public-market snapshot is about 7.85x. Calculated on a 2026-06-12 A+H combined basis, Hua Hong's current total P/S is about 15.3x, almost more than double those mature-node peers. The problem is that Hua Hong's 2025 gross margin was only 11.8%, clearly below UMC's 29.0% and GF's 24.9%, and not in the same tier as VIS's profit quality either. In other words, the market is giving Hua Hong a premium for being a "scarcer domestic-substitution platform," not for being "more profitable."

Whether this premium can hold for the long term depends on two questions. First, can Hua Hong truly lift low-teens gross margin? Second, will the A-share market's high premium for the company keep influencing perception of the company's overall valuation? As of 2026-06-12, the H-share price translated into USD was about USD 17.77, while the A-share price translated into USD was about USD 32.47, a premium of about 83%. Looking only at the H shares, Hua Hong is already not cheap. If the A shares are also treated as part of the company's total market capitalization, valuation rises another step. For new capital, the A+H structure is the easiest place for Hua Hong's valuation to create an illusion: the H shares look "only somewhat expensive," but once the A shares are consolidated, the whole company becomes clearly expensive.

Cash-Flow Look-Through

Net operating cash flow in 2025 was CNY 5.065 billion, while net profit attributable to the parent was CNY 377 million, making OCF/net profit extremely high. This looks more like the typical behavior of a heavy-asset manufacturer during a high-depreciation phase than evidence of accounting-profit fabrication: the income statement is under pressure, but cash flow is not equally ugly because depreciation is added back. The real issue is capital expenditure. In 2025 alone, cash paid for the purchase and construction of fixed assets, intangible assets, and other long-term assets reached CNY 12.931 billion, far above operating cash flow, so free cash flow remained clearly negative.

If 30%-40% of 2025 capex is roughly treated as maintenance capex, owner earnings are only about negative USD 2 million to USD 169 million, corresponding to a yield of about -0.04% to 0.46% on the current total market capitalization. This result has two meanings. First, what is least valuable at Hua Hong today is "current cash return distributable to shareholders," rather than "accounting profit." Second, future valuation cannot rest on the lazy assumption that "profit will naturally recover." It must rest on whether Fab 9 and Fab 5 can lift both gross margin and owner earnings. Before that happens, P/B has more explanatory power than P/E.

Absolute Valuation Scenarios

The following valuation separates operating variables from tolerable valuation multiples and is not investment advice. Because the company is in a heavy expansion phase, price ranges use the H-share primary quote translated into USD and anchor on 2025 year-end net asset value per share of USD 3.806. The conservative case is the buy range required when demanding a margin of safety, not "intrinsic value." The neutral case is the acceptable holding range when fundamentals have not deteriorated materially. The optimistic case corresponds to the upper edge where the market continues to grant a high premium.

Table note: The conservative/neutral/optimistic price ranges correspond to about USD 11.4-13.3, USD 16.0-18.3, and USD 20.9-23.6, respectively, all based on the H-share price translated into USD.

In plainer language, Hua Hong's problem today is "how much growth the current price has already bought in advance," rather than "whether there is growth." In the neutral scenario, the H-share price is already broadly inside the reasonable holding range, leaving no obvious margin of safety for new capital. If the A shares are also included, the company's overall valuation leans even more toward the market paying in advance for the next three years' story rather than pricing existing profits and cash flow.

Expectations Gap and Margin-of-Safety Review

The key expectation currently embedded in the market is that gross margin will keep moving up from the 13% line rather than falling back, and that Fab 9's peak capacity and Fab 5's new capacity can be absorbed by high-quality orders. The data most likely to create an expectations gap are gross margin, owner earnings, and management's comments on utilization and ASP, not revenue. If later reports show "revenue keeps growing, but gross margin still wanders around 12%-13%," the market will start questioning revenue quality.

The current price is at a clear premium to the conservative buy range, so the margin of safety is zero. The most fragile assumption in the three scenarios in section 7.3 is that "high utilization can offset new depreciation and keep pushing gross margin upward." If that assumption is discounted by 30%, the neutral range would be closer to USD 14-16 rather than the current USD 16-18.3. Another simple calculation says the H shares currently trade at trailing P/E above 500x and a static earnings yield of only around 0.2%. Under the condition of "zero profit growth over the next three years," returns would depend almost entirely on valuation not contracting, which itself proves that new buying lacks a margin of safety. This is a very typical "good factory, bad price" risk.

Margin-of-safety conclusion: none. For investors who already hold the H shares, this is more like a stock that can continue to be tracked, but cannot be justified by telling oneself it is "cheap." For new capital, waiting for a lower price or more solid margin recovery is the more reasonable discipline. The A shares are more aggressive; they are already clearly front-loading the future rather than simply having "no margin of safety."

Risk Analysis

The first and most practical risk for Hua Hong is mature-node price war. VIS management has publicly complained that Chinese peers have built "unnecessary capacity" in mature nodes, leading to intense price competition. SEMI also shows that China's 300mm equipment investment will remain high in 2026-2028. The probability of this risk is medium to high, and its impact is high. Observable indicators include whether the company's gross margin falls below 12% for two consecutive quarters, whether management again emphasizes "pricing" and "order selection," and whether peer quotations for mature nodes weaken. Once a price war truly begins, the transmission path is very direct: ASP falls, utilization may not fall immediately, but unit profitability is compressed first, and valuation multiples eventually contract as well.

The second risk is mismatch between expansion and depreciation. Fab 9's total investment is about USD 6.7 billion; more than USD 5.0 billion had been invested by the end of 2025, and a large cash outflow remains in 2026. The company is also advancing Fab 5 integration and the next phase of Fab 9. The probability of this risk is medium, and the impact is also high. Observable indicators are whether capex cash outflow exceeds expectations, whether operating cash flow can expand in step, and whether gross margin still sits in the low teens after Fab 9 reaches peak capacity. If new capacity arrives first while orders and ASP arrive later, Hua Hong will enter the awkward zone of "larger revenue but thinner profit," and the stock narrative will move from "platform expansion" to "capital consumption."

The third risk is structural constraint from geopolitics and equipment supply. Hua Hong has already discussed raising the proportion of domestic equipment, which indicates that management assumes access to certain external equipment may be constrained. The probability of this risk is medium, and the impact is medium to high. Observable indicators include whether the company's language on domestic equipment and materials procurement continues to strengthen, whether later project construction is delayed, and whether customers' localization strategies change direction. Its transmission is through slower line construction, yield ramp, and customer certification delaying revenue realization, rather than an immediate hit to profit.

The fourth risk is execution of related-party restructuring. The Fab 5 acquisition makes sense industrially and can indeed solve part of the horizontal-competition issue and add capacity, but it is still a related-party transaction. The probability of this risk is medium, and the impact is medium. The items to watch after completion are three practical things, not announcement language: whether customers and platforms are truly integrated, whether R&D and manufacturing resources are truly shared, and whether margins and cash flow improve as a result. If the transaction only brings more revenue without better gross margin and operating efficiency, its meaning for ordinary shareholders will be heavily discounted.

The fifth risk is the A/H valuation mismatch itself. The large gap between Hua Hong's H-share and A-share prices shows that different markets are assigning completely different risk premiums to the same company. If the A-share premium keeps lifting perception of the company's overall valuation, it can make the narrative hotter in the short term. If capital style shifts and the A-share premium falls, it can also feed back into the market's reference for reasonable H-share valuation. The probability of this risk is high, and the impact is medium. It does not harm the factory itself, but it directly influences stock returns.

Catalysts and Tracking Indicators

Positive catalysts are not mysterious. The two strongest are whether gross margin can move from 13% toward 15%-18% for several consecutive quarters, and whether management can prove over one to two quarters that Fab 9 peak capacity is a real ability to bring higher-quality revenue rather than an accounting number. If the Fab 5 transaction lands smoothly and quickly brings customer synergy on the 55/40nm platform, that will also be a marginal positive.

Negative catalysts are equally clear. First, revenue keeps growing while gross margin weakens. Second, operating cash flow is not bad, but capex keeps pushing free cash flow deep into negative territory. Third, peers begin to talk more openly about price cuts and order transfers. Fourth, the high A-share premium falls and compresses the overall valuation framework.

Table note: These thresholds are research-tracking disciplines, not official company targets.

Among these indicators, the most important are gross margin, owner earnings, and the H/A premium, not quarterly revenue. Revenue tells you whether demand exists. Gross margin tells you whether the orders are good orders. Owner earnings tell you whether fab expansion is creating value for shareholders. The H/A premium reminds you whether the market has started to become excessively excited. For a heavy-asset cyclical manufacturer like Hua Hong, the most damaging setup is always "orders look plentiful, but profit is not," rather than "orders are a bit lower."

Zen Horizon Synthesis

Over its history, Hua Hong has truly proved two capabilities. First, it can deepen specialty processes, and this has not been a one- or two-year sprint. From putting Hua Hong NEC assets into the listed platform, to the Hong Kong listing, the National Integrated Circuit Industry Investment Fund investment, the Fab 9 ramp, and the Fab 5 acquisition, Hua Hong has consistently revolved around platforms such as eNVM, power, analog power, logic, and RF. The path may not look glamorous, but it is highly consistent. Second, it can find its position between policy, cycles, and capital markets. Many manufacturers know how to build fabs but do not explain why they deserve long-term capital. Hua Hong is the opposite: it knows how to turn points such as "localization," "specialty processes," and "platform synergy" into a narrative the market can understand.

Yet its most successful historical phase also benefited from the era and the cycle. The high profits in 2021-2022 came from demand explosion, tight supply, and rising mature-node pricing. They did not grant the company a permanent license to 25%-30% gross margin. How much of those success factors remains today? Demand remains, specialty processes remain, and localization remains. But industry supply discipline is less attractive, Fab 9 depreciation is still there, and the next expansion phase has not stopped. Hua Hong's real advantage today is the "depth" of China-based specialty processes and relatively strong industrial-organization capability. Its real shortcoming is that margins and free cash flow have not yet caught up with the speed of valuation resetting.

The market is most likely to misjudge two things. First, to confuse "high utilization" with "high earnings quality." Utilization is indeed high, with an average of 106.1% in 2025, but gross margin still hovers around 13%. Second, to confuse "domestic-substitution platform" with "valuation can naturally be more than double mature-foundry peers." Platform scarcity is real, but platform value ultimately has to return to cash flow. Capital markets can give it patience for a while in advance, but cannot indefinitely bear the reality that owner earnings are close to zero.

The most important variable over the next year is whether gross margin can move away from the 13% platform. The most important variable over the next three years is whether, after Fab 9 reaches peak capacity, new capacity can turn into high-quality revenue and pull owner earnings back into positive territory. The most important variable over the next five years is whether Hua Hong can turn "specialty-process platform" from a domestic-substitution concept into an asset class within global mature nodes that can price steadily. Under what conditions would it become a better investment? The answer is direct: a lower price or a higher margin, not a bigger story. Under what conditions should the original view be overturned? If utilization falls, gross margin deteriorates continuously, free cash flow remains deeply negative, and Fab 5 synergies remain invisible, the company will retreat from "valuation resetting" back into a "capital-consuming cyclical fab."

Bull and Bear Cases

Bull case:

• Hua Hong has real accumulated capability in specialty-process platforms such as eNVM, power, analog power, and logic/RF, and its product mix is not pure power-device price-war exposure.

• Average capacity utilization reached 106.1% in 2025, showing that current expansion is not empty expansion and demand absorption has been validated by the market.

• Fab 5 brings about 38,000-40,000 wafers per month of 55/40nm 12-inch specialty-process capacity, helping fill platform, customer, and scale gaps.

• In the geopolitical environment, European and some U.S. customers are advancing "China for China," placing Hua Hong in a beneficiary position for localized manufacturing.

• Fab 9 reaching peak capacity in 2026 and the next phase advancing in 2027 give real room for the revenue platform to lift over 3-5 years.

Bear case:

• Gross margin remains around 13%, with margin recovery far slower than revenue recovery, showing that new capacity has not yet formed high-quality returns.

• Free cash flow was materially negative in 2025, owner earnings are very thin, and the current valuation is not friendly to cash returns.

• China's 300mm capital expenditure is still rising, future mature nodes face supply pressure, and price war is not a theoretical risk.

• Combined-basis valuation is already materially above mature-node peers such as UMC, VIS, and GF, while the company's margins and cash-flow quality are weaker.

• The A/H price gap is too large, and the high A-share premium can easily push perception of the whole company's valuation into an overheated zone.

Pre-mortem

• Scenario one: In 2027, China's mature nodes enter more visible price competition. Nexchip, SMIC, and other 12-inch mature-node supply keep coming online. To protect utilization, Hua Hong is forced to follow prices down, gross margin falls from 13%-15% back to 9%-10%, and the market reprices it from a "platform asset" to 2.5-3.0x P/B. The H-share price translated into USD falls from around 17.8 to USD 8-10, close to a halving. The key variable in this scenario is whether orders have profit, not whether there are orders.

• Scenario two: After Fab 9 reaches peak capacity, loading materializes more slowly than equipment and depreciation. Fab 5 integration also fails to bring immediate platform synergy. In 2026-2027, operating cash flow remains decent, but capex and owner earnings continue to deteriorate. The market loses patience and treats the high-multiple story stock again as a capital-consuming manufacturer. At that point, a share-price decline does not require demand collapse; it only requires capital markets to stop front-loading the future.

Final Research Conclusion

Hua Hong Semiconductor is worth studying and deserves long-term tracking because it is not an empty concept. It really does hold a rare technology platform and capacity platform in China's local specialty-process foundry market. Fab 9 is not a factory in a slide deck, and Fab 5 is not a related-party transaction without industrial meaning. The real question is how much of the future the market is already willing to pay for, not whether the company is good. By mid-2026, Hua Hong has been rewritten by the market from a "profit elasticity stock" into a "platform re-rating stock." That gives it more imagination, and also makes it easier to overvalue.

On the current H-share primary quote basis, the stock is already in the zone of "holdable, but uncomfortable for new capital." Fundamental repair is real: revenue, utilization, and product mix are all improving. This round of valuation uplift is not driven by sentiment alone. The point requiring more caution is that margins, owner earnings, and capital returns still lag the share-price narrative. For existing holders, the most reasonable action is to focus tracking on whether gross margin continues to rise and whether Fab 9/Fab 5 truly produce cash returns, rather than rushing to sell. For new money, waiting for a better price or a clearer margin step is the better discipline.

The conditions for changing the view are also clear. If gross margin can stabilize at 15%-18% over the next two to three quarters, and the gap between operating cash flow and capex begins to narrow materially, Hua Hong's valuation will move from "front-running" to "delivery." Conversely, if revenue continues to grow while margin does not move, or if mature-node price competition starts eroding profitability, the current valuation will become fragile. Hua Hong's fabs can be very valuable, provided the money ultimately left to shareholders also grows.

【Company Profile Scores】

• Fundamental quality: Medium

• Growth: Medium

• Moat: Medium

• Financial resilience: Medium

• Management credibility: Medium

• Valuation attractiveness: Low

• Risk level: High

• Suitable investor type: Cyclical

【Investment Rating】

• Rating: Hold

• One-sentence investment thesis: Platform value is real, but the current price has already capitalized margin and cash-flow recovery ahead of delivery.

• 【Ideal/Fair Buy Price】11.4-13.3 USD Basis: corresponds to 3.0-3.5x P/B, leaving a margin of safety for low-teens gross margin and expansion execution risk.

• Holdable price: 16.0-18.3 USD

• Clearly overvalued price: above 20.9 USD

• Current price classification: holdable

• Worth waiting for a better price: yes. New capital is better suited to wait for the H shares to return near USD 11.4-13.3, or to wait until gross margin clearly moves above 15% before paying a higher price. The opportunity cost of waiting is that it may miss a stretch of valuation continuation driven by Fab 9/Fab 5.

• Target holding period: 1-3 years

• Expected annualized return: conservative around -30%; neutral -4% to +3%; optimistic around +25%

• Maximum downside risk: if price war and depreciation pressure overlap from 2027, the H-share primary basis has 50%-55% drawdown risk

• Signals that trigger reassessment: gross margin below 12% for two consecutive quarters

• Average capacity utilization falls below 90%

• After Fab 9 reaches peak capacity, the gap between operating cash flow and capex does not narrow

• After Fab 5 integration, revenue rises but margin does not improve

• The A/H premium stays above 80% for a long period while the H shares also begin rising away from fundamentals

The following ranges are all USD prices per H share.

【Valuation Range】

• current: 17.77 (as of the 2026-06-12 close)

• bear (conservative, ideal buy zone): [11.4, 13.3]

• base (reasonable, acceptable holding zone): [16.0, 18.3]

• bull (optimistic, above the clearly overvalued line): [20.9, 23.6]

Key Data Table

Table note: 2025 cash flow and share-capital structure come from the A-share annual report and 2026Q1 report. Revenue, gross margin, and utilization prioritize the company's USD-based announcements and earnings briefing.

Research Uncertainties

• This search did not directly extract the original table for Q2 2025, so the "last four quarters" trend judgment mainly relies on primary nodes from 1Q25, 3Q25, 4Q25, and 1Q26. Directional judgment is credible, but quarterly-detail completeness is slightly weaker.

• Common sell-side wording in the market says "Fab 9 phase one will reach full mass production within two years." The primary records found in this search are more specific: peak capacity will be reached in 2026, and the next phase's main work will begin in 2027. The latter basis is prioritized below.

• The company has not officially split maintenance and expansion components within 2025 capex. Owner earnings are a research estimate, not a company-disclosed figure.

• The A-share annual report's "gross margin of principal operations" differs in scope from the company's USD quarterly gross margin. This report has explicitly chosen the latter as the primary valuation basis.

• Because the company is dual-listed in A shares and H shares with a very large price gap, the H-share primary quote range cannot be mechanically translated to the A shares. The A shares reflect more local liquidity and thematic premium.

Reference Sources

• Hua Hong Semiconductor 2025 annual report, 2026 first-quarter report, 2025 fourth-quarter results announcement, and Q1 2026 earnings-briefing record.

• Hua Hong Semiconductor A-share prospectus and 2026 report on issuing shares to acquire Huali Micro.

• Shanghai Stock Exchange, Hong Kong market pages, and China Money central parity exchange-rate data.

• UMC, VIS, and GlobalFoundries official investor relations and financial-report materials.

• SEMI, WSTS, and Reuters industry materials on global fab investment, mature-node competition, and supply-chain restructuring.

• Nexchip public disclosures and market snapshots, used to observe mainland China's 12-inch mature-node challengers.

Other Stocks Mentioned in the Report

• 00981.HK — mainland China's largest foundry, and Hua Hong's most important local scale reference.

• 2303.TW — a mature-node cash-return foundry, and a direct comparison for Hua Hong's earnings quality.

• 5347.TW — a specialty mature-process peer whose public comments on China price competition are highly useful as a reference.

• GFS.US — a representative of globalized mature and specialty-process foundry, used to compare multi-region capacity and valuation discipline.

• 688249.SHG — mainland China's third-largest foundry challenger, representing the expansion of local 12-inch mature-node supply.

• 2330.TW — the technology and capital-expenditure ceiling of the global foundry industry, used as an industry upper-bound reference.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.