Giant Network Draws Institutional Buying as Tongfu Microelectronics Faces a ¥979 Million Exit

Giant Network attracted ¥165 million in net institutional buying on July 29, while Tongfu Microelectronics faced ¥979 million in net selling from institutional seats. The contrast appeared in post-market trading disclosures summarized through the rsshub 36kr news feed.

The same disclosure showed institutions trading in 39 stocks. Sixteen recorded net institutional buying, while 23 recorded net selling. Star-Net Communication and Suzhou GMT Technology followed Giant Network among the strongest institutional purchases.

The numbers create a striking reversal. Tongfu Microelectronics had recorded ¥661 million in net buying from five institutional seats only weeks earlier. July 29 did not simply extend that enthusiasm. It produced a much larger flow in the opposite direction.

That reversal matters more than the daily rankings alone. China’s public trading disclosures reveal which exchange-designated seats handled the largest transactions during unusually active sessions. They do not identify every investor or explain each trade’s purpose.

The result is a useful but incomplete signal. Institutional desks favored a profitable gaming company while reducing exposure to several technology and semiconductor names. Investors now must decide whether this was temporary risk control or an early valuation reset.

What the July 29 Institutional Trading Data Showed

The clearest signal was not Giant Network’s purchase alone, but the breadth and size of institutional selling elsewhere.

According to the original market disclosure, institutional seats appeared in the public trading records of 39 stocks on July 29. Sixteen finished with net institutional buying, compared with 23 showing net selling.

Giant Network ranked first among the net purchases at ¥165 million. Star-Net Communication followed at ¥160 million, while Suzhou GMT Technology received ¥90.73 million.

Those three companies sit in different parts of the technology market. Giant Network develops and operates online games. Star-Net Communication supplies networking and communications products. Suzhou GMT Technology manufactures precision electronic components.

The selling side carried much larger numbers. Tongfu Microelectronics recorded ¥979 million in net institutional selling. Unisplendour Corporation followed at ¥483 million, while Honghe Electronic Materials recorded ¥473 million.

Tongfu Microelectronics provides semiconductor assembly and testing services. Unisplendour sells networking, computing, and digital infrastructure products. Honghe Electronic Materials produces electronic-grade fiberglass fabric used in printed circuit boards.

The ranking therefore did not show institutions leaving every technology segment. It showed selective buying alongside concentrated exits from several hardware-related names.

That distinction is essential. A daily list with more sellers than buyers does not establish a broad institutional retreat. The relevant question is where the largest amounts accumulated.

The three leading net purchases totaled about ¥415.73 million. Tongfu Microelectronics alone recorded more than twice that amount in reported net selling.

Adding Unisplendour and Honghe Electronic Materials brings the top three reported exits to ¥1.935 billion. That imbalance explains why Tongfu became the day’s central story despite Giant Network leading the buying list.

A public trading list is also narrower than the entire market. Exchange rules publish selected information when securities meet conditions tied to price movements, turnover, or abnormal trading.

The Shenzhen Stock Exchange says it publishes market data and public transaction information every trading day. Its disclosure rules determine which unusually active securities receive detailed listings.

This means the 39 stocks were not a random sample of all listed companies. They were stocks whose trading qualified for additional visibility.

An institutional seat is similarly not a complete investor identity. It indicates that an exchange-classified institutional channel appeared among the disclosed buyers or sellers. It does not reveal whether one portfolio, several funds, or hedged accounts drove the total.

Net buying subtracts reported sales from reported purchases for the relevant seats. It describes activity within that disclosed set, not the institution’s total ownership change across every account.

These limitations do not make the numbers meaningless. They define what the numbers can support.

July 29 produced verified evidence of a sharp split within the disclosed trading sample. It did not prove that institutions had collectively abandoned Chinese semiconductor companies or permanently embraced gaming stocks.

The useful takeaway is narrower. Giant Network drew meaningful institutional demand, while Tongfu Microelectronics experienced an unusually large and visible institutional exit.

Why Giant Network Attracted the Strongest Buying

Giant Network entered July 29 with earnings momentum and recent investor interest, giving buyers a clearer near-term case than many speculative technology names.

Giant Network reported first-quarter revenue of ¥2.329 billion for 2026, according to its company announcement summarized in the quarterly results. That represented a 221.70 percent increase from the corresponding period.

Net profit attributable to shareholders reached ¥1.080 billion. The reported year-over-year increase was 210.58 percent.

Those figures offer one plausible foundation for July 29’s ¥165 million net institutional purchase. Investors were not buying only a distant promise. They were looking at a company that had reported substantial earnings growth.

The trading signal also followed another highly active session earlier in July. On July 15, Giant Network reached its daily price limit and recorded ¥695 million in net buying across the disclosed leading seats.

Northbound trading channels accounted for most of that earlier net purchase, according to the July 15 record. That session showed broader investor enthusiasm before the July 29 institutional ranking.

The two disclosures should not be merged into one continuous position. Different seats and different market conditions produced each result. Still, the earlier activity establishes that July 29 did not emerge from an inactive market.

Giant Network also offers investors a different earnings profile from semiconductor assembly companies. Game publishers can benefit when successful titles produce high-margin revenue without requiring equivalent increases in physical production capacity.

That operating model does not eliminate risk. Game revenue can depend heavily on a limited number of titles, user retention, content schedules, and regulatory approvals.

However, strong reported earnings give institutions a concrete figure to test. Investors can compare future revenue and profit against the first-quarter surge rather than relying entirely on long-range industry demand.

The company’s recent investments also place it near the market’s artificial intelligence narrative. Giant Network participated in financing for VAST, a developer working on 3D models and world-model technology.

It also joined a strategic financing round for Lightwheel Intelligence. These transactions offer exposure to AI development, although their direct contribution to Giant Network’s current profit remains unclear.

Investors should avoid treating those investments as proof that Giant Network has become an AI company. Its disclosed earnings and gaming operations remain more important to the immediate financial case.

This is where the institutional preference becomes understandable. Buyers could combine an established gaming business, strong recent results, and optional exposure to AI-related ventures.

Star-Net Communication and Suzhou GMT Technology received smaller but still significant net purchases. Their presence confirms that institutions did not avoid hardware universally.

Star-Net Communication’s ¥160 million inflow was only ¥5 million below Giant Network’s total. That small gap suggests the buying side was not a one-company event.

Suzhou GMT Technology’s ¥90.73 million inflow added another electronic-component company to the leaders. Institutional buyers were therefore selecting among technology names rather than drawing a simple boundary between software and hardware.

Still, Giant Network held the strongest visible position. Its ¥165 million net purchase arrived alongside recent earnings growth that investors could measure immediately.

That combination separates Giant Network from a purely thematic trade. Institutions appeared willing to pay for current operating momentum while keeping exposure to future technology narratives.

The next test will be whether later financial statements preserve the first quarter’s growth. A single quarter can reflect a favorable comparison period, title timing, or revenue recognition patterns.

If subsequent reports show stable player spending and durable profit, July 29’s buying will look more fundamental. If growth falls sharply, the institutional purchase will resemble a short-term momentum trade.

The Tongfu Microelectronics Reversal Was the Real Market Signal

Tongfu Microelectronics shifted from heavy institutional buying to much larger selling within three weeks, turning one daily ranking into a test of semiconductor expectations.

On July 9, Tongfu Microelectronics reached its daily price limit after generating ¥11.339 billion in turnover. The stock’s reported turnover rate was 10.76 percent.

Five institutional seats recorded ¥661 million in net buying that day, according to the July 9 trading data. The Shenzhen Stock Connect channel bought ¥1.379 billion and sold ¥484 million.

By July 29, the reported institutional balance had reversed to ¥979 million in net selling. The change between those two visible institutional readings was ¥1.64 billion.

That calculation does not mean the same institutions bought and then sold every share. Public seat data cannot confirm continuous ownership at the account level.

It does establish a dramatic change in disclosed flow. Institutional channels were aggressive net buyers during one unusually active session and dominant net sellers during another.

This reversal is the article’s primary tension. Semiconductor demand can remain favorable while investors simultaneously reduce exposure to a stock whose valuation or positioning has moved too far.

Tongfu Microelectronics operates in outsourced semiconductor assembly and testing. These businesses package fabricated chips and verify that they meet performance and reliability requirements.

Assembly and testing companies can benefit from growing chip volumes and greater packaging complexity. They also face heavy capital needs, customer concentration, pricing pressure, and cyclical utilization rates.

Artificial intelligence demand adds another layer. Advanced processors require sophisticated packaging, but the economic benefits do not reach every supplier equally.

Investors must distinguish between broad AI enthusiasm and the revenue captured by a specific company. A strong industry forecast cannot substitute for verified orders, improving margins, and rising capacity utilization.

The July 9 purchase suggested institutions were willing to embrace the upside. The July 29 exit suggests at least some disclosed desks were unwilling to hold that exposure without limit.

Profit-taking is one possible explanation. Risk reduction is another. Portfolio rebalancing, hedging activity, or a response to company-specific information can also produce large net selling.

The public data do not identify the motive. No responsible interpretation should claim that institutions discovered a hidden operational problem solely from the trading list.

Yet the size of the exit cannot be dismissed. Tongfu’s ¥979 million net selling exceeded Giant Network’s reported institutional buying by ¥814 million.

It also exceeded the combined net buying reported for Giant Network, Star-Net Communication, and Suzhou GMT Technology by roughly ¥563 million.

That concentration makes Tongfu the clearest pressure point. It shows how quickly institutional positioning can change when a crowded technology narrative meets valuation discipline.

The earlier July activity provides an important historical reference. Investors had already demonstrated significant demand for Tongfu, with both institutional and northbound channels appearing prominently.

July 29 therefore was not simple neglect. It was active selling after active buying.

For North American readers, this resembles the pattern often seen around semiconductor momentum trades globally. Investors may agree with a long-term computing trend while disagreeing sharply about near-term expectations embedded in individual stocks.

The public data cannot reveal the valuation model behind the trades. They can reveal that institutional conviction was less stable than the July 9 purchase initially suggested.

That instability matters for anyone treating a leaderboard appearance as a durable endorsement. Institutional buying can reverse before a company publishes another major financial report.

The July 29 result should therefore be read as a positioning event. It raises questions about expectations, but it does not answer them.

What the rsshub 36kr Signal Cannot Tell Investors

The rsshub 36kr alert captures a real exchange-derived event, but it cannot establish who traded, why they traded, or what happens next.

RSSHub distributes structured feeds from public websites, while 36Kr publishes short business and technology news updates. In this case, the feed carried a concise summary of the July 29 institutional rankings.

That delivery method improves speed. It does not expand the underlying evidence.

The source report states how many stocks showed institutional activity and identifies the leading net purchases and sales. It does not provide a complete account-level history for the institutions involved.

This distinction prevents three common analytical mistakes.

First, an institutional seat is not synonymous with a single long-only mutual fund. Different professional investors can use institutional channels, and their objectives can vary.

One desk may be building a long-term position. Another may be reducing risk after a rapid gain. A third may be trading around derivatives or related holdings.

Second, net selling does not establish a negative fundamental judgment. Portfolio managers sell for liquidity, position limits, index changes, redemptions, taxes, and many other reasons.

The July 29 data contain no official statement from the selling institutions. Assigning a single motive would turn an observed transaction into speculation.

Third, a public trading list is not the full market. It highlights securities that met exchange disclosure conditions, usually following unusual price or trading activity.

The Shenzhen exchange’s 2026 trading rules provide the regulatory framework for transactions and public market information. These rules support transparency without exposing every investor’s complete portfolio.

The disclosure also should not be confused with a corporate announcement. Giant Network did not announce that institutions had endorsed its strategy. Tongfu Microelectronics did not announce that institutions had rejected its outlook.

These were secondary-market transactions reported through exchange mechanisms. They changed ownership among market participants but did not directly add cash to either company.

The strongest conclusion must remain proportional to the evidence. On July 29, disclosed institutional seats bought more Giant Network shares than they sold. They sold substantially more Tongfu shares than they bought.

Anything beyond that requires additional information.

Investors should also resist sector-level overreach. Unisplendour and Honghe Electronic Materials joined Tongfu among the largest institutional net sales, creating a visible hardware cluster.

However, Star-Net Communication and Suzhou GMT Technology appeared among the largest net purchases. That mixed result contradicts a simple “institutions sold hardware” narrative.

The better interpretation is selective repricing. Institutional desks separated companies according to earnings visibility, recent performance, valuation, liquidity, and portfolio exposure.

Giant Network’s first-quarter figures offered a direct earnings catalyst. Tongfu’s rapid shift from heavy buying to heavier selling highlighted the fragility of positioning around semiconductor expectations.

Unisplendour’s ¥483 million exit and Honghe’s ¥473 million exit strengthened the risk-reduction signal. Neither amount alone proves a deterioration in those companies’ operations.

Company disclosures remain the proper place to verify operational changes. Exchange rules require listed businesses to publish material information fairly and promptly.

The Shenzhen exchange’s fair disclosure guidance says significant information should reach investors equally. It also prohibits selective disclosure of undisclosed material information to favored parties.

That framework gives readers a practical hierarchy of evidence. Audited statements and formal announcements should carry more weight than inferred motives from seat activity.

Daily trading data still have value. They reveal where professional trading became large enough to appear in an exchange’s selected public record.

They also expose changes in market behavior before quarterly ownership reports become available. That timeliness makes them useful for generating questions.

They are weaker at answering those questions. The rsshub 36kr signal showed the direction and magnitude of selected flows, not the investment thesis behind them.

Investors should use it as an alert to investigate company filings, earnings quality, valuation, and future disclosures. It should not function as an automatic buy or sell instruction.

Three Signals That Will Test the Institutional Split

The next one to three months should show whether July 29 marked a durable rotation or a temporary adjustment after volatile trading.

The first signal is Giant Network’s next financial update. Investors need to see whether the company can extend the revenue and profit momentum reported for the first quarter.

The first-quarter comparison was unusually strong. Revenue increased 221.70 percent, while attributable net profit rose 210.58 percent.

If later results preserve a substantial portion of that momentum, July 29’s ¥165 million institutional purchase will gain fundamental support. It would suggest buyers were responding to earnings durability.

If growth falls sharply or profitability weakens, the interpretation changes. The purchase would look more dependent on recent price momentum and optimistic expectations.

Investors should focus on operating performance rather than promotional language. Revenue composition, title contribution, user spending, and profit quality will matter more than broad references to gaming or AI.

The second signal is Tongfu Microelectronics’ next earnings and capacity disclosure. The central question is whether operating results justify the enthusiasm visible earlier in July.

Useful indicators include revenue growth, gross margin, capital expenditure, capacity utilization, and customer demand. These measures can show whether semiconductor packaging demand is improving the company’s economics.

If margins and utilization strengthen, the ¥979 million exit will look more like profit-taking or portfolio rebalancing. The long-term operating case would remain intact despite volatile ownership.

If those measures disappoint, July 29 will appear more consequential. The selling would have occurred before weaker expectations became visible in later public results.

Investors should remain careful even then. Similar timing does not prove that sellers possessed undisclosed information.

Fair disclosure requirements exist precisely because material information should reach all investors at the same time. Evidence of an operational change must come from formal filings or verified statements.

The third signal is whether later public trading lists repeat the same sector pattern. One day can reflect temporary liquidity needs, but repeated flows create a stronger positioning trend.

Watch whether Giant Network continues attracting institutional purchases after its next volatile sessions. Also watch whether Tongfu, Unisplendour, and Honghe repeatedly appear with net selling.

A repeated pattern would strengthen the rotation thesis. It would suggest institutions were persistently favoring companies with visible earnings over hardware names carrying demanding expectations.

A reversal would weaken that thesis. Renewed institutional buying in Tongfu or Unisplendour would show that July 29 captured a temporary adjustment rather than a lasting sector judgment.

Star-Net Communication and Suzhou GMT Technology will provide useful controls. Both received institutional buying while other technology names experienced large exits.

If those companies continue receiving demand, the data will reinforce a stock-selection narrative. If they join the selling list, broader technology risk reduction becomes more plausible.

Readers should also compare disclosed trading with company announcements. A large flow deserves more attention when it coincides with verified changes in earnings, contracts, regulation, or capital spending.

Without that confirmation, the flow remains a market behavior signal. It measures action, not motive.

The contrast between Giant Network and Tongfu Microelectronics is therefore more informative than either number alone. One company combined strong reported earnings with the day’s largest institutional purchase.

The other moved from ¥661 million in institutional buying on July 9 to ¥979 million in selling on July 29. That swing challenged the idea that professional enthusiasm creates a stable floor.

The broader leaderboard added weight to the caution. Twenty-three of 39 disclosed stocks showed net institutional selling, and the largest exits greatly exceeded the leading purchases.

Yet the data stopped short of a universal retreat. Institutions still bought selected gaming, networking, and electronic-component companies.

For readers following the rsshub 36kr update, the right next step is to track the three tests in order. Start with Giant Network’s earnings durability, then examine Tongfu’s operating economics, and finally compare future institutional disclosures.

Do not treat one leaderboard as a portfolio instruction. Use it to identify where expectations and professional positioning have separated most sharply.

Will Giant Network convert its first-quarter surge into sustained profit, or will Tongfu’s next results make the July exit look premature? The answer will determine whether July 29 marked a genuine rotation or only another volatile trading day.