This set of strategies has netted Hefei a total of one trillion yuan.

On July 27, 2026, Changxin Technology was listed on the Sci-Tech Innovation Board, with its opening price surging by more than 450%, and its intraday total market value once exceeded 3 trillion yuan.

Calculated based on the closing market value of the day, the approximately 33% equity held by the Hefei state-owned capital system has a book value of more than 1 trillion yuan.

What is the concept of 1 trillion yuan? The general public budget revenue of Hefei in 2025 was less than 100 billion yuan.

One investment, over a period of ten years, has a book return equivalent to the city's ten years of fiscal revenue.

Public opinion quickly boiled over, with labels such as "the best venture capital in China" and "the most daring city to take risks" spreading everywhere.

In these noisy narratives, a word that people love to mention is "reckless gamble".

In fact, this word has an implied connotation, suggesting that Hefei's move is a bet-style adventure and a lucky victory. But I think these remarks are too "sour". You have to wonder, why did Hefei dare to allocate 13.5 billion yuan in 2016 to bet on a field that was almost completely blank in China at that time? Why did Hefei not withdraw its capital during the eight years when Changxin Memory accumulated losses of more than 36 billion yuan? Why is it Hefei, rather than other cities with stronger financial strength, that has accomplished this?

In my opinion, the real value of the Changxin case lies not in proving that "local governments can make a lot of money through investment", but in showing how a complete set of institutional arrangements can turn "long-termism" from a slogan into an executable operating system.

This system consists of at least three core components, namely the risk resolution mechanism, industrial chain integration capability, and counter-cyclical decision-making authorization. Beyond these three components, there is still an unsolved exit problem, which even relates to whether this trillion-yuan floating profit can finally be cashed in.

Fault-tolerance mechanism, changing the behavior of decision-makers?

Why does Hefei "dare" to invest? In the vast majority of local governments, state-owned capital investment decision-makers face "asymmetric risks". For example, if the investment is successful, the income belongs to the public, and individuals may not get direct returns; if the investment fails, once losses occur, audits, inspections and accountability will follow, and their careers may be ruined as a result.

Under this structure, any rational official will tend not to invest, or only invest in projects that "cannot go wrong". And projects that "cannot go wrong" usually mean mature technical routes, verified business models, and closed risk windows. To put it bluntly, they are fields that others have already explored and paid the tuition for.

This "chasing the rise and selling the fall" investment logic runs counter to the "betting in uncertainty" required to cultivate emerging industries.

Hefei has broken this incentive structure. According to public information, Hefei has set an overall risk tolerance of up to 40% for its angel investment funds, allowing a maximum of 100% loss for a single project. More critically, a practical list of "due diligence exemption" has been introduced as supporting measures. As long as the investment decision has gone through the prescribed due diligence procedures, information disclosure and collective decision-making processes, the relevant responsible persons will not be held excessively accountable even if the project eventually fails.

This system makes an institutional separation between the objective fact that "a project may fail" and the organizational response that "someone must take responsibility for it". The significance is that the state-owned capital investment team has a reassurance: what they need to do is to make full preparations for the project, rather than guaranteeing the success of every project, which is impossible in the first place.

Therefore, this also explains why Hefei did not withdraw but continued to increase investment even when Changxin Memory suffered consecutive losses for eight years, with accumulated losses exceeding 30 billion yuan? The answer is not "courage", but the system.

When the personal career risk is underwritten by the system, decision-makers can shift their attention from "what will happen to me if I lose money" to "whether this project objectively still has opportunities".

This point has direct reference value for the current nationwide implementation of "patient capital".

Many local governments have issued administrative measures for government investment funds, and have also mentioned "fault tolerance", but most of them stay at the level of principled statements. For example, terms such as "fully disclose risks" and "complete procedure performance" lack quantitative standards and clear division of responsibilities.

In practice, audit departments still tend to judge based on the profit and loss of a single project.

As a result, fault tolerance is written in documents, but accountability is still implemented in practice.

Hefei's experience shows that to make the fault tolerance mechanism really work, at least three elements are required: a quantitative range of allowable loss rate, a positive exemption list, and inversion of the burden of proof, which means that the supervising party needs to prove that the investment team has faults, instead of the investment team proving its own innocence.

The 16 new regulations on state-owned assets issued by Shanghai in April 2026 have clearly defined the direction of "combining annual assessment with long-cycle assessment" and "not simply taking the profit and loss of a single project or a single year as the assessment basis".

The direction is correct.

But Hefei's practice shows that from "correct direction" to "implementable", there is a whole set of implementation rules in between.

This is also the space where Shanghai can make efforts in the next step.

Invest in the missing links of the industrial chain

The fault tolerance mechanism solves the problem of "daring to invest", but it cannot solve the problem of where to invest.

Hefei does not invest in "good projects", but invests in "missing projects". The difference of one word leads to completely different logics.

"Good projects" is from the perspective of financial investment, with good track, good team, rapid growth and reasonable valuation. But it cannot answer the core proposition of local government industrial investment: what does this project mean to the local industrial ecosystem after it lands? What upstream and downstream industries can it drive? Which short board on the industrial chain can it make up for?

If you just invest in a project with good financial performance that has no connection with local industries, it is essentially just a financial investment, which has nothing to do with the long-term construction of urban competitiveness.

Hefei's practice is to first clarify what industries the city wants to develop, then find the most missing, most difficult and most critical links on the industrial chain of this industry, and then use state-owned capital to fill the "market failure" in that link.

Take the integrated circuit industry as an example. Hefei clearly proposed to build an "IC Capital" around 2013. The background of this strategic decision is that home appliances, flat panel displays and automobiles are the three pillar industries of Hefei, and these three industries all encountered the same bottleneck in the process of transformation and upgrading - lack of chips.

If the chip link cannot realize local supporting facilities, the competitiveness of the entire manufacturing cluster will be subject to others.

After clarifying "what we want", the next step is to figure out "what we are missing".

In the field of memory chips, China's independent production capacity is almost zero, and 96% of the global market is monopolized by three companies: Samsung, SK Hynix and Micron.

The DRAM link is the largest and most fatal short board on Hefei's industrial chain.

Changxin Memory was built to target this short board.

Its value does not lie first in how much money it can make in the future, but in whether its existence can drive the entire industrial chain to gather in Hefei.

Facts have proved that this logic is valid. Around the "chain leader" Changxin, a number of enterprises such as Cambricon, Tongfu Microelectronics and Pexton Technology have settled in Hefei one after another. At present, Hefei has gathered more than 450 upstream and downstream integrated circuit enterprises, covering the whole chain of design, manufacturing, packaging and testing, materials and equipment.

In 2025, the output value of Hefei's integrated circuit industry exceeded 1514 billion yuan, an increase of about 7.4 times compared with about 180 billion yuan in 2016.

In this logic, the role of state-owned capital has undergone a qualitative change.

It is not "selecting projects", but "organizing industrial chains". Investing in Changxin drives the agglomeration of hundreds of enterprises; the agglomeration of hundreds of enterprises in turn provides lower cost and higher efficiency supply chain supporting facilities for Changxin.

This is a positive cycle.

The initial investment of state-owned capital leverages the multiplier effect of the entire industrial cluster.

Another key node of this investment method is that Hefei always adheres to the positioning of "industrial organizer" rather than "passive investor" in the investment process. From helping Changxin jointly invest to purchase the patent package of Canadian Wi-LAN to bypass technical barriers, to establishing special leading groups at the municipal and provincial levels to provide full-process services for project construction and financing, Hefei state-owned capital has always been deeply embedded in every key link of industrial organization.

This is more complicated than simply writing a check, but it is this deep embedding that makes "accurate investment" possible.

Counter-cyclical investment, where does the institutional authorization come from

However, there is a third question, that is, when to invest?

Changxin is exactly such an example. Hefei invested in it when the industry was at its worst.

2023 was the cold winter of the global memory industry. The three giants Samsung, SK Hynix and Micron all sharply reduced their capacity utilization rates and cut capital expenditures to "survive by cutting their losses".

At that time, Changxin Memory had not yet made a profit, and its attributable net loss in 2023 was as high as 16.34 billion yuan.

According to common sense, the most rational choice at this time is to shrink, wait and see, and wait for the spring to come.

Hefei chose to increase investment against the trend, supporting Changxin to continuously increase R&D investment and ramp up production capacity.

The production line utilization rate increased from 85.45% to 94.63%. This decision was later proved to be the key winning move. When the demand for AI computing power detonated the memory chip market in 2025, Changxin was the enterprise that had prepared sufficient production capacity.

But at that time in 2023, what made the decision-maker dare to make such a choice?

This is related to the institutional authorization of counter-cyclical investment. In fully market-oriented investment institutions, counter-cyclical investment requires the investment committee to trust the fund manager. But in the state-owned capital system, counter-cyclical investment faces double difficulties. First, the book losses are expanding, and the pressure from audits and public opinion is increasing; second, there is a high degree of uncertainty in the judgment of industry prospects, and no one can provide conclusive evidence for "increasing positions now".

If decision-makers bear the risk of being held accountable for the loss of a single project, then in any normal organizational rationality, they will not choose to increase positions at this time.

Shrinking is the safest choice, and safety often overrides everything in the state-owned capital system.

The root cause why Hefei can make counter-cyclical decisions can still be traced back to the first dimension: the fault tolerance mechanism.

It is precisely because of the underwriting of the quantitative allowable loss rate and the institutional guarantee of due diligence exemption that decision-makers can act according to industrial logic rather than personal risk aversion logic at the moment when everyone is in fear.

There is also a detail here. It is reported that in its industrial investment, Hefei adopts a set of "two-line parallel, independent decision-making" mechanism, that is, the state-owned capital platform conducts project research and judgment and investment decision-making according to market-oriented investment logic, and government departments conduct investment attraction assessment and landing guarantee according to industrial policy logic. The two lines do not interfere with each other, and are finally coordinated at the municipal level.

This design is very exquisite. It not only ensures the professionalism of investment decision-making, not kidnapped by short-term administrative considerations; but also ensures the strategic nature of industrial layout, not led by the financial return of a single project.

At the critical moment of counter-cyclical, this "two-line parallel" structure allows the long-term judgment based on industrial logic to penetrate the short-term fear based on financial indicators.

For cities seeking to improve the management system of state-owned capital funds, Hefei's practice suggests a specific direction: in strong cyclical industries such as integrated circuits and biomedicine, "counter-cyclical investment trigger clauses" can be set in advance. For example, when the prosperity index of a sub-sector is lower than a certain threshold, the special counter-cyclical investment quota is automatically activated, and a more loose assessment standard is applied.

In this way, "whether to invest counter-cyclically" is changed from a case-by-case decision that requires repeated games to a set of institutional arrangements that can be agreed in advance.

So why did Hefei state-owned capital make a huge profit of one trillion yuan?

In my opinion, the fault tolerance mechanism ensures "daring to invest", the industrial chain logic ensures "accurate investment", and the counter-cyclical institutional authorization ensures "investing at the worst time".

The combination of these three components forms a complete closed loop from front-end decision-making to back-end management.

It is the result of institutional design, not a product of luck.

But this closed loop is still missing the last piece of the puzzle, that is, the maturity of the exit mechanism.

Only when you can "exit smoothly" after "investing smoothly", and can "re-invest smoothly" after "exiting smoothly", can this system complete its institutional construction truly.

At present, this step is still on the way.

For cities that are learning the Hefei model, the real enlightenment is not "how many billions to invest in a project", but whether they can establish a complete set of institutional infrastructure that makes long-termism operable, such as quantifiable fault tolerance, in-depth cultivation of the industrial chain, authorization for counter-cyclical investment, and smooth exit channels.

Missing any link, what you learn is only superficial.

For Hefei itself, the listing of Changxin is not the end. It is not only the moment of realization after ten years of long-distance running, but also the starting point of a more difficult journey ahead. How to convert the trillion-yuan floating profit into sustainable fiscal return and industrial re-investment capacity will determine whether the "Hefei model" is just a legend that can be told repeatedly, or a set of institutional paradigm that can be replicated and promoted.

We should continue to pay attention to this issue.

This article is from the WeChat official account "Dongzhen Business Review", author: Dongzhen Business Review, published with authorization from 36Kr.