Venture Capital Seeking Anchor: Behind the Frequent Emergence of 100-Million-Yuan Angel Rounds

"If we still insist on conducting three months of in-depth due diligence, the TS (Term Sheet) will be out of the queue for consideration." Recently, a founding partner of a Beijing-based equity investment institution frankly admitted that he has felt this pervasive anxious sentiment spreading across the venture capital circle more and more frequently.

Several projects that this partner has contacted this year not only no longer allow independent in-depth due diligence by investment institutions, but some even set conditions during financing: the intended capital must be split into two rounds of investment at different valuations, otherwise the investment institutions cannot obtain the share of this round.

Three to five years ago, if an angel round project put forward such requirements, it would most likely be blacklisted by partners of top-tier VCs (Venture Capital). But now, this has become the new normal in cutting-edge tracks such as embodied intelligence, general large models, and commercial aerospace.

Along with the changes in the capital side structure, the early-stage investment logic of the primary market is also being reconstructed. The "tens of millions of RMB angel round" that was once regarded as industry news now has its threshold raised to the level of hundreds of millions or even billions of RMB. But on the other side of the coin, behind this carnival of high valuations and fast pace for top-tier targets, the financing of most ordinary projects is still struggling on the brink of life and death, and the once most capital-intensive growth stages of Series B and Series C are still facing severe risk of capital flow interruption.

Money Flowing Upstream

Lin Da (pseudonym), a partner at a top FA institution in Beijing, has been extremely busy recently. His main task is to help several large model and embodied intelligence enterprises "matchmake" early-stage shares within a short period of time.

"The key word for current early-stage financing has changed from 'whether you can get the money' to 'how much 'big' money you can get'", Lin Da sighed to the Economic Observer reporter. For several projects he has handled, the decision-making time for investment institutions has been extremely compressed, and even seller market terms such as "no in-depth due diligence allowed, two rounds of parallel advancement at different valuations" have emerged.

Cao Jishan, founding partner of Cloud Road Capital, also said, "Two years ago, we priced angel round hard technology projects with a very clear, almost uncontroversial standardized system, the core of which was designed around the 'certainty already in hand'. Now when pricing top projects in top tracks such as artificial intelligence and embodied intelligence, the traditional fundamental anchor point has failed."

Macro data echoes this perception to a certain extent. The "2026 China Private Equity Investment Industry Observation Report for the First Three Quarters" (hereinafter referred to as the "Observation Report") released by ZERONE shows that the investment and financing activity in the primary market continued the recovery trend in the first three quarters of 2026, and capital is tilting to the early stage with unprecedented intensity. The data shows that in the first three quarters of this year, the total proportion of seed, angel round and Series A financing events reached as high as 73.9%, among which Series A is still the single financing round with the highest proportion, but the proportion of seed and angel rounds has rebounded significantly compared with 2024, reaching 34.8%.

Capital has not decreased, it has only become more concentrated and generous.

From the perspective of capital flow, new quality productive forces have become the main force attracting capital. In the first three quarters of 2026, the proportion of strategic emerging industries in the overall investment and financing events has continued to rise to 84.7%. In terms of specific track distribution, artificial intelligence leads by a large margin with a financing scale of 2035 billion RMB, followed by the robot track which attracted 87.5 billion RMB.

It is worth noting that within these popular tracks, the investment logic is also undergoing subtle shifts. Taking embodied intelligence as an example, the focus of capital is shifting from the early concept of building ontology to the basic layer of building brains, skins and hands. Startups focusing on electronic skin, force control joints and multimodal perception such as Lige Wannwu, Gewu Tongzhi and Mogan Technology have obtained the capital volume that used to be available only in Series A or even Series B rounds at the seed and angel stages. Early-stage medical investment is also shifting from virtual to real, switching from telling stories to counting certificates and clinical trials.

Capital Landscape Transformation

At the end of October 2026, an enterprise named HyperVolve attracted much attention in the primary market. This company, registered at the end of July that year and with actual operation of only more than two months, won hundreds of millions of RMB in angel round financing without releasing any products, with zero revenue and zero customers. The list of investors behind it is extremely luxurious: not only led by BlueRun Ventures, but also includes SenseTime Guoxiang Capital, Legend Star and Shuimu Tsinghua Alumni Seed Fund.

It is not a unique case. Hefei Qianyuan Quantum announced an over 100 million RMB angel round jointly invested by provincial, municipal and district state-owned assets only one month after its establishment. Lige Wannwu, founded by the former head of foundation models at ExDynamics, received a bet from HSG right after its first month of establishment.

How come some companies that haven't even finished building their official websites can be valued at hundreds of millions of RMB at the angel round?

Many interviewed investors pointed out to the Economic Observer reporter that it is no longer the traditional performance-based financing. What investors buy is not financial reports and commercialization progress, but a long-term option about what the next-generation foundational model or platform will look like.

This end-state reversed pricing model includes three layers of premiums that are difficult to be quantified by conventional financial models. The first is team premium, that is, the irreplaceability of the core team, for example, the experience of ten-thousand-card training and the ability to build industry-level evaluation systems. The second is path premium: capital is shifting from the application layer which is easily wiped out by the replacement of underlying models, and starts to heavily invest in foundational models with cross-scenario migration capabilities. The last is time point premium: compared with similar overseas projects that often have valuations of billions of US dollars, the scarce early top-tier targets in China seem cheap at valuations of several hundred million RMB.

What supports this set of high premium logic is the change of the primary market capital structure.

The "Observation Report" released by ZERONE shows that in the first three quarters of 2026, the investment activity of institutional LPs (Limited Partners) rebounded sharply, with more than 10,000 subscribed commitments and the total subscribed amount reaching 1.02 trillion RMB. In this 1.02 trillion RMB of flowing capital, guiding funds and state-owned enterprises are still the largest source of investment, with subscribed investment of 560.7 billion RMB, accounting for 54.7%.

A more significant change is the rise of industrial capital. In the first three quarters, industrial investors invested 190.7 billion RMB, and their proportion jumped to 18.6%, becoming an important supporting force second only to state-owned capital. This figure has increased by 42% significantly compared with the same period last year.

"The players at the table have changed the rules now," said an investment director of a dual-currency fund. In the past, state-owned capital often played the role of a pure investor, but now, state-owned capital LPs come to invest with industrial maps. Where the project can land and which local industrial chain it can complete, its priority is often placed before pure financial returns. This also forces many GPs (Fund Managers) to revise their fundraising and investment pitches, and whether they can help state-owned capital LPs complete the investment attraction landing assessment has become the core competitiveness of institutions.

On the other hand, the trend of industrial capital (CVC) entering the market to seize early-stage shares is becoming more and more intense. Large manufacturers such as Xiaomi, Alibaba, China Mobile and Ant Group no longer wait until the situation becomes clear in Series A and Series B rounds to make moves as they did in the past.

In addition, the entry of long-term capital such as insurance funds, as well as the extension of the state-owned capital assessment window on the policy side and the implementation of the due diligence compliance exemption mechanism, objectively provide a patient capital environment for these hard technology early-stage projects that often take ten years to hone a sword and are difficult to see cash flow in the short term.

Behind the Carnival

However, in this super angel round carnival focusing on top-tier targets, the other side of the market appears extremely deserted.

Lin Da, the aforementioned partner of the FA institution, frankly admitted that the current market is very fragmented. For projects with the same target financing amount of 100 million RMB, if they are invested in embodied intelligence, commercial aerospace or solid-state batteries, there will be institutions queuing up to give money; but if they are invested in model innovation, or ordinary projects whose technology is not on the independent controllable list, the valuation may be directly cut by 70%.

According to data from CVSource by ChinaVenture, there were 974 domestic early-stage investment events in the third quarter of 2026, with a total disclosed scale of about 60.8 billion RMB. Although the average single financing amount has been raised to more than 62 million RMB, the median single financing amount only stays at 20 million RMB.

The problems faced by the mid-section of the venture capital industry are also severe.

The front-end seed and angel rounds are crowded with state-owned capital, industrial capital and top-tier VCs; the back-end Pre-IPO stage is supported by national large funds. However, the growth stages of Series B and Series C, which used to have the most active capital transactions, have experienced a capital gap due to the ebb of traditional private VCs and the obstruction of exit channels.

This gap may bring new crises. Many non-head startups obtained 10 to 20 million RMB of startup capital in the early stage relying on the track dividend, but when they really enter the stage of product polishing and need large-scale verification of commercialization closed loop, they find that the follow-up capital cannot keep up, and finally can only be trapped helplessly in the "Death Valley".

The carnival of high valuations does not last forever. Many investors revealed to the Economic Observer reporter that cases of flat rounds or even down rounds have begun to appear in the market recently. Some projects that were forcibly issued at a valuation of 2 billion RMB in the angel round, when entering the Series A round, subsequent investors are only willing to accept a post-money valuation of 1.8 billion RMB or even lower.

The general consensus in the industry is that once the proportion of flat rounds (valuation remains flat) or down rounds (valuation is lowered) in Series A financing exceeds 30%, it marks that the upward cycle of early-stage valuations has basically ended. The high premium supported by forward narrative is bound to face systematic re-pricing.

This article is from the WeChat Official Account "Economic Observer", author: Hong Xiaotang, published with authorization from 36Kr.