Unitree Robotics is going public.
With a market cap of 60.9 billion yuan and a P/E ratio of 219 times, it aims to raise over 6 billion yuan, oversubscribing by 45%. The online subscription was overwhelming, resulting in a pitifully low allotment rate.
We previously discussed that Meituan is Unitree's largest external shareholder ('Wang Xing Becomes Wang Xingxing's Largest External Shareholder'), and Meituan is likely to reap huge profits.
Today, IT桔子 talks about those who act as both referees and players behind the scenes – the brokerages.
What you might not know is that for this Unitree IPO, at least 12 brokerages quietly "laid the groundwork" by investing before the listing. During the subscription phase, another 34 brokerages queued up to participate in the IPO.
And CITIC Securities, the lead underwriter for the listing, alone made money four times over on this single project.
12 Brokerages: They Bought In Long Ago
Unitree had 46 shareholders before going public. Peeling back the layers of these shareholders reveals the shadows of at least 12 brokerages.
How did they get in? Mostly not through direct purchases, but by taking detours.
For example, the fund Jinshi Growth is under CITIC Securities.
Before Unitree's listing, Jinshi Growth held 4.15%, making it the fifth-largest shareholder.
But the Jinshi Growth fund doesn't just contain CITIC's own money – Hualong Securities invested a bit, Huaan Securities invested a bit, and Hongta Securities also invested a bit. Everyone pools money together, hands it to CITIC's fund to invest – this is the so-called LP (Limited Partner) model.
You chip in a little, I chip in a little, we invest in star projects together, and share the profits upon listing.
Besides the Jinshi Growth route, there are other paths: Soochow Securities holds indirectly through Shenzhen Capital Group's fund, Guotai Junan and Haitong hold indirectly through the Shanghai Science and Technology Innovation Fund, CICC holds indirectly through CICC Capital...
With all these twists and turns, the core logic is one: the alternative investment subsidiaries of brokerages give money to top-tier PE funds, using them as a conduit to acquire Unitree shares.
This kind of operation isn't new in the industry, but it's quite rare to see 12 brokerages simultaneously appear on a single project's shareholder list.
CITIC Securities: One Dish, Eaten Four Times
If other brokerages are hitching a ride, then CITIC Securities is "driving the car, selling the tickets, and collecting the toll."
On the Unitree project, CITIC did four things simultaneously:
First, it was an early-stage investor. Through two Pre-IPO investments by Jinshi Growth and China Securities Investment, it invested roughly over 33 million yuan. Calculated at the issue price, this investment is now worth nearly 500 million yuan, with a return rate of around 1300%.
Second, it was a co-investor. The STAR Market has a rule: the sponsor must buy a bit too. CITIC, through CITIC Investment, bought over 800,000 shares for 122 million yuan, locked up for 24 months.
Third, it was the sponsor and lead underwriter. Unitree raised over 6 billion yuan this time. The sponsor fee plus underwriting fee add up to roughly 145 million yuan. This is CITIC's most direct "service fee" income.
Fourth, it was also the manager of the employee strategic placement asset management plan. Unitree's senior executives and core employees participated in the strategic placement, and their money is managed by CITIC's asset management plan, locked up for 36 months. Founder Wang Xingxing himself put in 15 million yuan.
Let's do the math: equity investment earned a 1300% return, co-investment locked for 24 months waiting for the lock-up to lift, sponsor/underwriting fees securely pocketed 145 million yuan, plus managing the employees' strategic placement funds.
One dish, eaten four times.
This model has a benefit: CITIC's own money is also locked in, so it has the incentive to do the project well and not just muddle through. But the question is: when you are both the largest institutional shareholder and the pricing sponsor, how much independence do you have left?
There's no standard answer to this question, but it's worth thinking about.
34 Brokerages Queued Up for the IPO
On the subscription day, it was another scene of excitement.
For Unitree's offline IPO subscription, 34 brokerages and their asset management subsidiaries participated. From top players like CITIC, Huatai, Guotai Junan, and Haitong, to mid-sized ones like Guoyuan, Zheshang, and Caitong, even foreign brokerages' China subsidiaries like JPMorgan and Goldman Sachs showed up.
These 34 firms together were allotted over 650,000 shares, amounting to roughly 98 million yuan.
Interestingly, some of these brokerages already indirectly held Unitree shares in the primary market through LPs, and now directly participated in the IPO through asset management products in the secondary market. That means adding another layer on top of their existing position.
Why are they still scrambling to buy even with a 219 times P/E ratio?
Simply put, brokerages judge that Unitree still has room to rise after listing. Offline IPO participation is not charity; it's based on calculated accounts.
These institutions believe that even at such a high P/E, there's still short-term profit to be made.
Three Things Worth Remembering for Ordinary Investors
First, "taking detours" is the norm. Most small and medium-sized brokerages don't hold shares directly but hold them indirectly through the LP model. This means a stable profit-sharing relationship has formed between the brokerage system and top-tier PE funds. You provide the money, I provide the project, and we share the profits.
Second, primary + secondary: brokerages' total exposure is larger than you think. 12 brokerages hold shares in the primary market, 34 participate in the secondary market IPO, with overlapping lists. Some brokerages both indirectly hold shares in the primary market and directly participate in the IPO in the secondary market. Combined, the brokerage system's actual exposure on Unitree is much larger than the direct shareholding ratio seen in the prospectus.
Third, at 219 times P/E, everyone is in the same boat. Unitree's estimated non-GAAP net profit for 2025 is about 590 million yuan. At 219 times P/E, the market cap is over 60 billion yuan. Is this valuation high? High. But the brokerage system, as a whole, is both the largest camp of institutional shareholders and the most core participant in pricing.
When the sponsor's own funds, employee strategic placement funds, and underwriting income are all locked in the same ultra-high valuation project, this is no longer just "aligned interests" – this is shared risk.
Is shared risk good? Yes. But the premise is that Unitree's post-listing performance can deliver.
In the first half of 2026, Unitree expects revenue of 1.05 to 1.13 billion yuan. Whether it can maintain high growth for the full year is a crucial test for the humanoid robot sector, moving from storytelling to performance scrutiny.
The brokerages have bet on Unitree with more than just underwriting fees.
In the first half of 2026, Unitree expects revenue of 1.05 to 1.13 billion yuan.
Whether it can maintain high growth for the full year is the key verification window for the humanoid robot sector to move from concept to performance. 12 brokerages laying groundwork in the primary market, 34 brokerages participating in the IPO in the secondary market, CITIC Securities eating one dish four times – under the high valuation of 219 times P/E, this is both aligned interests and shared risk.
This article is from the WeChat public account "IT桔子" (ID: itjuzi521), author: Judy
