China’s STAR Market Profit Surged 437.6%, but One Sector Drove Most of the Gain

China’s STAR Market reported a 437.6% increase in first-half net profit, lifting the technology-focused board’s earnings to 144.887 billion yuan. Revenue reached 1.01 trillion yuan, up 38.6% from the same period in 2025.

Those figures, released by the Shanghai Stock Exchange on August 30, describe one of the strongest reporting periods since the board opened in 2019. They also create a striking contrast with the wider Shanghai market, where net profit increased 17.6%.

The headline deserves attention, but it does not describe a uniform recovery across every technology company. Semiconductor businesses generated most of the reported profit, while 36 companies in the market’s growth tier remained collectively unprofitable.

That concentration is the central tension. The results show real commercialization across chips, artificial intelligence, and biotechnology. However, they also show how one capital-intensive industry can dominate an aggregate market statistic.

The STAR Market’s 437.6% Profit Jump in Context

The most important change is not the percentage alone, but the gap between profit growth and revenue growth.

The exchange summary says STAR Market companies generated 1.01 trillion yuan in combined first-half revenue. That was 38.6% higher than a year earlier.

Combined net profit reached 144.887 billion yuan, an increase of 437.6%. First-half profit also exceeded the board’s result for all of 2025, according to the exchange.

Revenue growth near 39% already represents a sharp expansion. Yet profit grew more than eleven times as quickly as revenue. That difference indicates that sales growth alone cannot explain the result.

The percentage also starts from a relatively low comparison base. Applying the reported growth rate to the current figure implies approximately 26.9 billion yuan of net profit in the first half of 2025.

That arithmetic helps explain why the percentage is so large. A recovery from a depressed base can produce an extraordinary growth rate even when the current profit level is less extreme.

The result still matters. A low base does not manufacture 144.887 billion yuan in current earnings. It simply changes how readers should interpret the 437.6% figure.

The wider Shanghai market provides another useful reference. Its 2,318 listed companies produced 26.22 trillion yuan in first-half revenue, up 6.3%. Net profit reached 2.82 trillion yuan, up 17.6%.

Profit excluding nonrecurring items rose 17.2% to 2.69 trillion yuan. The exchange described that as the strongest growth rate since 2022.

Against that background, STAR Market revenue grew more than six times as quickly as the wider market. Its reported profit growth was about twenty-five times the wider market rate.

Manufacturing also performed better than the broad exchange. Shanghai-listed manufacturers increased revenue by 12.8% and profit by 40.3%, according to the same disclosure.

The STAR Market therefore sits at the leading edge of a broader manufacturing recovery. It is not an isolated pocket of improvement, but its earnings acceleration remains exceptional.

Profitability measures support part of the story. The median gross margin among STAR Market companies was 36.8% during the period. Gross margin measures the revenue left after direct production costs.

STAR Market research and development spending reached 104.4 billion yuan, up 14.6%. The median ratio of research spending to revenue was 12.6%.

That combination matters because higher earnings did not coincide with a retreat from research. The board increased R&D spending while reporting much faster profit growth.

However, an aggregate median does not establish that every company improved. It also does not show whether gains came from recurring operations, investment income, asset sales, subsidies, or accounting changes.

The available exchange summary provides a market-level result, not a complete attribution for every company. Individual financial statements remain necessary for judging earnings quality.

The first conclusion is therefore narrower than the headline. STAR Market companies delivered a major earnings rebound, supported by fast revenue growth and continued research spending. The distribution of that rebound requires closer examination.

Semiconductors Accounted for Most of the Profit

The profit surge was principally a semiconductor story, not an equal advance across the technology board.

Shanghai-listed integrated-circuit companies generated 125.267 billion yuan in first-half net profit, according to the exchange. Their combined profit increased 660.3% from the previous year.

That figure equals roughly 86.5% of the STAR Market’s reported 144.887 billion yuan in net profit. The two totals do not necessarily cover identical company sets, but their scale reveals the sector’s influence.

Semiconductor companies also reported a median gross margin of 33.5%. The exchange attributed the improvement to stronger domestic supply chains, higher utilization, equipment adoption, and advances in packaging.

ChangXin Technology was the largest example cited in the review. The exchange says the memory-chip company generated 77.6 billion yuan in first-half net profit after moving from a loss a year earlier.

That single result represents more than half of the STAR Market’s aggregate profit. It also explains a substantial share of the board’s year-over-year reversal.

ChangXin’s latest LPDDR6 memory product and improving capacity utilization were among the operating developments highlighted by the exchange. Capacity utilization describes how much installed production capacity a manufacturer actually uses.

Higher utilization can produce an outsized profit effect in semiconductor manufacturing. Chip fabrication requires large upfront investment, while many factory costs remain fixed across production levels.

When factories process more wafers, those fixed costs spread across a larger volume. Revenue can rise while unit costs fall, creating profit growth that outpaces sales.

This operating leverage offers a plausible mechanism for the sector’s results. It also creates risk because the same mechanism works in reverse when demand weakens or prices fall.

Semiconductor cycles are sensitive to inventories, capital spending, product pricing, and customer concentration. One strong half does not remove those characteristics.

Other companies added breadth to the semiconductor story. Semiconductor Manufacturing International Corporation and Hua Hong Semiconductor each reported record quarterly revenue, according to the exchange review.

Equipment makers also advanced. Advanced Micro-Fabrication Equipment says its twelve-inch etching systems now support several critical steps for devices at three nanometers and below.

Piotech’s thin-film deposition equipment had supported an accumulated production volume of approximately 600 million wafers at customer facilities. Deposition equipment places controlled material layers onto a wafer during chip fabrication.

Materials suppliers reported growth as well. China Shipbuilding Industry Corporation’s specialty-gas unit nearly tripled revenue from tungsten hexafluoride, a material used in semiconductor manufacturing.

Xi’an ESWIN Material Technology exceeded monthly capacity of one million twelve-inch silicon wafers. SJ Semiconductor also moved a micro-bump-based 3D integrated-circuit platform into volume production.

These examples suggest the improvement reached several layers of the supply chain. Fabrication, equipment, materials, and advanced packaging all contributed to the exchange’s semiconductor narrative.

Still, investors should separate operational evidence from promotional framing. The exchange presents these developments as evidence of stronger domestic technology capacity, but it does not publish segment-level attribution for the entire 125.267 billion yuan.

The sector total can also mask major differences among memory producers, foundries, design companies, and equipment suppliers. Their economics, customer bases, and exposure to price cycles differ substantially.

The 660.3% growth rate therefore requires the same caution as the board-level number. Part of the gain reflects current business expansion, while part reflects the prior period’s weak or negative earnings base.

The semiconductor contribution makes the overall result more understandable. It also makes the headline more concentrated than it first appears.

Artificial Intelligence Added Revenue, but Not the Main Profit Surge

Artificial intelligence strengthened the growth case, although the disclosed numbers do not make it the primary source of market profit.

Four representative STAR Market computing-chip companies, including Hygon Information Technology and Cambricon Technologies, produced 18.155 billion yuan in combined revenue. That was 82.2% higher year over year.

The exchange says their main products had been adapted for leading domestic large language models. It also highlighted work on supernode systems and clusters containing thousands of accelerator cards.

A supernode links multiple processors through high-speed connections so they can operate more like one computing system. The approach targets the communication bottlenecks that appear during large-model training and inference.

Those developments matter because China’s AI market faces constraints across chips, networking, memory, software compatibility, and electrical capacity. Selling an accelerator alone does not resolve the entire computing problem.

Compatibility with domestic models can reduce one adoption barrier. Larger clusters can also support customers that need more aggregate computing capacity than a single server provides.

However, the exchange disclosed combined revenue for the four representative chip companies, not their aggregate net profit. That distinction limits how directly they can explain the board’s 144.887 billion yuan profit total.

Upstream suppliers showed clearer earnings gains. Shengyi Electronics increased net profit by 109%, while Founder Technology reported a 232% increase. China Jushi increased net profit by 74%.

These companies supply infrastructure and materials used beyond a single AI product category. Their results suggest AI investment can transmit demand through circuit boards, manufacturing capacity, and composite materials.

Software provided another signal. Kingsoft Office’s enterprise subscription revenue from AI office products increased by more than 60% for a sixth consecutive quarter, according to the exchange.

That is relevant because enterprise subscriptions test whether AI features can support recurring revenue. Hardware demand can move with construction cycles, while subscriptions provide a different commercialization path.

Telecommunications companies also began packaging token consumption into market offerings. A token is a small unit of text or data processed by an AI model.

This approach turns model usage into a measurable service input. Yet the disclosed summary does not provide token revenue, margins, customer retention, or the share generated by external buyers.

Embodied AI added another growth example. The exchange says Unitree Robotics ranked first globally in humanoid-robot shipments and increased first-half revenue by almost 50%.

That claim illustrates commercial momentum, but it lacks shipment totals and an independent market-share source in the disclosure. It should be treated as the exchange’s reported assessment.

The broader pattern is more defensible. AI demand appears to be supporting revenue across processors, circuit boards, office software, telecommunications services, and robotics.

What remains unproven is whether those businesses can generate semiconductor-scale profits without continued heavy investment. AI infrastructure requires spending on factories, data centers, networking, and energy.

Model services also face falling usage prices and intense competition. Growing token volumes do not automatically produce higher margins when providers lower prices or absorb infrastructure costs.

Developers and enterprise buyers should watch deployment economics, not only benchmark performance. Hardware availability matters, but software compatibility and total operating cost often determine whether a system reaches production.

For knowledge workers, the signal is similarly practical. Sustained subscription growth matters more than the number of AI features announced because it indicates recurring customer adoption.

Teams evaluating AI products can preserve earnings releases, technical claims, and deployment notes in a searchable knowledge base. That makes it easier to compare later disclosures with earlier promises.

The AI segment supports the STAR Market growth narrative. It does not replace the more concentrated explanation supplied by semiconductor profits.

Biotechnology Shows a Different Commercialization Path

Biotechnology contributed a second route to better earnings through drug approvals, licensing agreements, and international demand.

Shanghai-listed companies secured approvals for seven Class 1 innovative drugs during the first half, according to the exchange. China classifies these medicines as new drugs that have not previously been marketed domestically or internationally.

One cited product was iza-bren, Baili Biopharmaceutical’s dual-antibody drug conjugate. The exchange described it as the first approved dual-antibody ADC globally.

An antibody-drug conjugate links a targeted antibody with a therapeutic payload. The design seeks to deliver treatment more precisely to cells carrying selected biological markers.

Approvals can move a biotechnology company from research spending toward product revenue. They do not guarantee adoption because commercialization still depends on clinical positioning, pricing, reimbursement, manufacturing, and physician acceptance.

Licensing supplied another source of value. Hengrui Pharmaceuticals and RemeGen each signed an overseas licensing agreement with a potential total value exceeding $5 billion during the period.

Potential transaction values require careful interpretation. They usually combine upfront payments with development, regulatory, and sales milestones that might never become payable.

Dizal Pharmaceutical received a $600 million upfront payment in a licensing transaction, according to the exchange. An upfront payment offers stronger near-term evidence than a headline maximum because it is not contingent on later milestones.

These agreements reflect a recognizable biotechnology financing cycle. A company develops an asset, licenses rights to another business, receives cash, and reinvests part of that cash into its research pipeline.

That model can release the value of intellectual property before a product achieves full global commercialization. It can also create uneven earnings because large licensing payments arrive at irregular intervals.

WuXi AppTec produced more than 10 billion yuan in first-half net profit for the first time. Its outstanding orders reached 66.4 billion yuan at the end of June, up 25%.

Outstanding orders provide visibility into contracted demand, although they are not the same as recognized revenue. Timing, cancellations, project scope, and customer decisions can affect conversion.

United Imaging offered a medical-equipment example. Its overall market share increased by 1.7 percentage points, while overseas business exceeded one-quarter of revenue.

The company’s overseas gross margin improved by seven percentage points. That combination suggests international expansion did not rely solely on lower-margin sales.

Biotechnology and medical technology therefore differ from the semiconductor mechanism. Chip profits can respond sharply to factory utilization, pricing, and production scale.

Drug developers can produce large gains through approvals or licensing payments. Research service providers depend more on order conversion, while equipment companies depend on installations and service relationships.

Combining these businesses in one market statistic can obscure those differences. A profit increase driven by memory manufacturing does not carry the same durability as one supported by recurring service orders.

The reverse is also true. A milestone payment can lift one reporting period without establishing a repeatable earnings base.

The board’s sector diversity provides some protection from a single commercial model. It does not eliminate concentration when one industry supplies most aggregate profit.

Readers should therefore track the composition of earnings, not only the total. Recurring product sales, service revenue, milestone payments, and manufacturing rebounds deserve different valuation assumptions.

The biotechnology results reinforce the wider commercialization theme. Years of research spending are producing approvals, licenses, orders, and equipment sales.

They also reinforce the need for company-level analysis. Aggregate growth cannot show which earnings will recur in the next reporting period.

What the 437.6% Figure Does Not Show

The headline says profits recovered, but it does not prove that profitability became broad, recurring, or evenly distributed.

The clearest counterpoint comes from the STAR Market growth tier. Its 36 companies increased combined revenue by 29.1%, but they still recorded a net loss.

Their aggregate loss narrowed by 62.3%. That is meaningful progress, yet loss reduction is not the same as positive net income.

The growth tier was designed for companies that meet specific technology and development criteria while remaining in earlier commercial stages. Their results expose the gap between revenue expansion and sustainable earnings.

A company can grow sales while remaining unprofitable because research, production, customer acquisition, and financing costs rise at the same time. That pattern is common among young biotechnology and hardware companies.

The contrast is important. The entire board reported 144.887 billion yuan in profit, while one recognized group within it remained collectively in the red.

Concentration also affects the comparison base. If a large semiconductor producer moved from a substantial loss to a large profit, the percentage change for the board would rise dramatically.

That does not invalidate the outcome. It means the growth rate measures a reversal as much as continuing expansion.

The exchange’s disclosure provides several checks on quality. Median gross margin remained 36.8%, research spending rose, and the broader Shanghai market generated stronger operating cash flow.

Shanghai-listed real-economy companies produced 1.52 trillion yuan in operating cash flow, up 35.1%. That cash flow covered net profit by 1.2 times.

However, those cash-flow figures apply to the broader group of Shanghai-listed real-economy companies. They should not be treated as a standalone STAR Market cash-conversion ratio.

The same boundary applies to overseas revenue. More than 1,050 Shanghai-listed real-economy companies generated 3.20 trillion yuan abroad, up 22.8% for a third consecutive year.

This provides useful market context, but it does not isolate the overseas exposure of STAR Market companies. The board could have a different geographic mix.

Investors also need to distinguish net profit from profit excluding nonrecurring items. The exchange published both measures for the broader Shanghai market but only headline net profit in its STAR Market summary.

That omission prevents a clean market-level comparison between reported and underlying STAR Market earnings. Individual filings can fill the gap, but the aggregate announcement cannot.

Capital intensity creates another uncertainty. Emerging-industry companies across Shanghai spent 406.6 billion yuan acquiring long-term assets, up 4.4%.

Capacity expansion can support future revenue. It can also produce depreciation and underused factories if demand, pricing, or product transitions move against producers.

Policy support is part of the environment as well. Shanghai processed 68 refinancing applications after revised measures took effect, representing 127.2 billion yuan in proposed financing.

The number of accepted applications increased 48%, while proposed financing rose 39%. Active capital formation can accelerate research and production, but financing access does not guarantee commercial returns.

Mergers and acquisitions increased too. Shanghai-listed companies announced 370 restructuring activities, including 32 major transactions worth more than 350 billion yuan.

Consolidation can improve scale or supply-chain control. It can also introduce integration risk, valuation risk, and acquired earnings that complicate organic growth comparisons.

The broader earnings review also reports that STAR Market exchange-traded funds exceeded 364 billion yuan in assets by the end of August. That was approximately 20% higher year over year.

ETF growth can broaden market access, but fund inflows do not validate company fundamentals. Prices and earnings can diverge when investors respond to policy, liquidity, or thematic demand.

These limitations lead to a disciplined interpretation. The first-half result confirms a large improvement in reported earnings. It does not establish that every sector reached the same stage of maturity.

It also does not prove that the 437.6% rate can continue. Repeating that growth would require an increasingly large absolute increase from a much higher base.

Three Signals That Matter More Than the Next Headline

The next test is whether STAR Market earnings become broader, more recurring, and better supported by cash.

The first signal is semiconductor profit concentration in the next reporting period. Investors should compare the integrated-circuit sector’s net profit with the board’s aggregate result.

A declining concentration ratio, alongside stable total profit, would strengthen the case for a broad technology recovery. Rising concentration would show that the headline still depends heavily on chips.

Within semiconductors, utilization, gross margin, and inventory deserve particular attention. Those indicators can reveal whether earnings come from durable demand or a temporary cycle.

The second signal is whether the 36 growth-tier companies cross from narrower losses into aggregate profitability. Their first-half revenue growth and 62.3% loss reduction provide a measurable starting point.

Positive aggregate profit would show that earlier-stage companies are converting commercial growth into earnings. A renewed widening of losses would weaken the board’s breadth story.

Company-level cash flow matters here. Revenue growth funded by receivables, inventory, or repeated financing carries different implications from growth that generates operating cash.

The third signal is the quality of commercialization across AI and biotechnology. For AI, watch subscription retention, hardware deployment, and revenue that follows completed installations.

For biotechnology, separate upfront licensing payments from contingent milestones. Also track whether approved drugs generate recurring sales and whether research-service orders convert into recognized revenue.

These signals address the main weakness in the current headline. They test composition and durability rather than asking whether reported profit increased again.

The market earnings data already show a board moving beyond research-only narratives. Chips are shipping, drugs are receiving approvals, and enterprise AI subscriptions are growing.

Yet commercialization does not follow one timeline. A memory factory, an AI software provider, and a clinical-stage drug developer face different cost structures and failure points.

That difference matters to developers and enterprise buyers as well as investors. A supplier’s financial capacity affects product support, infrastructure expansion, hiring, and long-term availability.

Technology teams should preserve technical disclosures alongside financial results instead of treating them as separate information streams. A searchable workflow can connect product claims with later revenue, margin, and deployment evidence.

The STAR Market’s first-half numbers deserve recognition as a major earnings reversal. Revenue increased rapidly, research spending continued, and several industries produced tangible commercial results.

The more consequential question is now measurable: can the board retain its current profit level while reducing its dependence on a single semiconductor rebound?

Watch the next filings for sector concentration, growth-tier profitability, and operating cash conversion. Those three signals will show whether 437.6% marked a durable transition or an exceptional comparison period.