How XtalPi Became China's First AI Pharma Stock: A Journey of Innovation and IPO Success

Economic Observer Reporter Zhang LingOn June 13, XtalPi (02228.HK) was listed on the Hong Kong Stock Exchange, becoming the first AI pharmaceutical stock in China. As the first technology company to be listed under the HKEX's Chapter 18C rules, XtalPi is also the first stock under Chapter 18C.

Three Hong Kong political and business leaders spoke at the listing ceremony of XtalPi, they are Charles Tang, Chairman of the Hong Kong Exchanges and Clearing Limited (HKEX), Paul Chan, Financial Secretary of Hong Kong, and Sun Dong, Secretary for Innovation, Technology and Industry of Hong Kong. Charles Tang told several founders of XtalPi that it is uncommon to have three leaders speak, which reflects their recognition of the first stock under Chapter 18C.

At the listing ceremony, Wen Shuhao, co-founder and chairman of XtalPi, said that XtalPi is building a new infrastructure for future industry R&D with AI and robotics technology. After going public, it will continue its business philosophy of "helping others is helping oneself." He encouraged more technology innovation companies to go public by utilizing Hong Kong's Listing Rule 18C, jointly fostering trillion-dollar enterprises in the AI era.

Before its IPO, XtalPi completed a total of 8 rounds of financing, with investors including institutions and companies such as Sequoia, Mirae Asset, Google, Tencent, and China Life. The company has raised a cumulative total of $732 million, ranking first in the total financing amount among global AI pharmaceutical companies.

In this IPO, XtalPi Holdings Limited priced its shares at HK$5.28 each, with a global offering of 187 million shares, raising a total of approximately HK$1 billion. Among this, the company introduced eight cornerstone investors who collectively subscribed to over HK$337 million worth of shares.

On its first day of listing, XtalPi opened 2% higher and surged up to 24.62% during the session to HKD 6.58 per share. At the close, XtalPi was trading at HKD 5.8 per share, up 9.85%, with a total market value of HKD 19.8 billion.

"Double" First Stock

In March 2023, the Hong Kong Stock Exchange added "Chapter 18C" to the Main Board Listing Rules. This chapter imposes higher requirements on technology attributes and mainly targets Specialist Technology companies, allowing tech companies without revenue or profit to be listed in Hong Kong. This has, to a certain extent, lowered the listing threshold for Specialist Technology companies.

The 18C rule applies to five major industries: next-generation information technology, advanced hardware, advanced materials, new energy and energy conservation and environmental protection, and new food and agricultural technology. Different from the 18A rule, which previously facilitated the listing of multiple biopharmaceutical companies, unprofitable companies seeking to list under the 18C rule must secure investment of 5% or more of their shares from "experienced independent investors."

In this IPO, XtalPi plans to use 75% of the proceeds to enhance its R&D and solution-providing capabilities, 15% to strengthen its commercialization capabilities both in China and internationally, and 10% for working capital and general corporate purposes.

From 2021 to 2023, XtalPi's revenue was 62.799 million yuan, 133 million yuan, and 174 million yuan, respectively. Its R&D expenditures were 213 million yuan, 359 million yuan, and 480 million yuan, respectively, with net losses of 2.137 billion yuan, 1.439 billion yuan, and 1.906 billion yuan, respectively.

In June 2023, Black Sesame Technologies, a developer of autonomous driving computing chips, became the first technology company to apply for an IPO under the 18C rule. However, its listing application lapsed in early January 2024 after failing to pass the Hong Kong Stock Exchange's listing hearing within six months. In March 2024, Black Sesame Technologies submitted its IPO application again and successfully passed the hearing on June 12. In November 2023, XtalPi became the second technology company to apply for an IPO under the 18C rule but managed to go public ahead of Black Sesame Technologies.

In China's AI pharmaceuticals industry, XtalPi is also the second company to push for an IPO, similarly taking a step ahead to successfully go public — in June 2023, Insilico Medicine filed its prospectus, planning to list on the main board of the Hong Kong Stock Exchange; in January 2024, Insilico Medicine’s prospectus expired; in March 2024, Insilico Medicine submitted its listing application for the second time.

Over the past year, the AI pharmaceuticals track in China has seen both successes and setbacks. While some companies in the same field have faced layoffs or even bankruptcy liquidation, others have been closely watching XtalPi and Insilico Medicine’s push towards IPOs. Previously, Insilico Medicine was considered a top contender to become China's first AI pharmaceutical stock, but when its prospectus expired, it sparked concerns within the industry about the future of AI-driven drug development. "XtalPi’s listing is an excellent signal for the market, peers, and ecosystem partners. Especially for our peers, I believe it will boost their confidence, which is very important," said Ma Jian, co-founder and CEO of XtalPi Holdings Limited, in an interview with the Economic Observer. As a representative technology platform company applying artificial intelligence and robotics in fields like chemistry, pharmaceuticals, and materials, XtalPi’s successful listing on the Hong Kong Stock Exchange Main Board is, in every aspect, a highly fulfilling achievement.

Three People, Ten Years

The three founders of XtalPi—Shuhao Wen, Jian Ma, and Lipo Lai—are classmates from MIT and often exchanged ideas together in the past. In 2014, cloud-based computing power began to grow exponentially, and the trio realized this would make high-precision and rapid large-scale computing a reality.

The following year, they founded XtalPi, leveraging first-principles calculations based on quantum physics and artificial intelligence capabilities to transform the design and discovery of drugs and new materials at a speed and scale surpassing traditional alternative methods.

In XtalPi's development history, Pfizer has been a significant partner. In 2016, Pfizer held a global crystal structure prediction competition, in which XtalPi participated and provided accurate predictions, earning Pfizer’s recognition. In 2021, XtalPi was commissioned by Pfizer to participate in the development of the oral COVID-19 drug Paxlovid. By combining predictive algorithms with experimental validation, XtalPi helped Pfizer confirm the drug’s crystal structure in just six weeks. These two events have been widely discussed within the industry.

Ma Jian said that these two things were repeatedly mentioned, not necessarily because they were the biggest issues, but because they occurred at the most critical moment for XtalPi.

Around Thanksgiving in 2016, Wen Shuhao took his colleagues to Pfizer's R&D center in the United States, while Ma Jian and other colleagues stayed in China. When the results were announced, it was around 3 a.m. in China. He remotely joined the competition, and after hearing the applause following the announcement of the results, the weight on his mind finally lifted. After that competition, XtalPi reached a cooperation agreement with Pfizer.

"This is our commercialization process from zero to one," Ma Jian said. Enterprises must first solve the problem of survival; no matter how good the technology is, it won't work if no one is willing to pay for it.

After building a strong reputation, XtalPi established an artificial intelligence research and development center, various drug discovery platforms, and smart robotic wet labs. It launched the brands "XtalPi Drug Discovery" and "XtalPi Intelligent Automation," promoting the further development of AI in new drug and new materials fields.

XtalPi and Pfizer have maintained a stable and continuous business partnership. Pfizer routinely sends drugs at a certain research stage to XtalPi for crystal-related studies. In October 2020, Pfizer delivered a new molecule, which XtalPi's team did not pay particular attention to since Pfizer had provided several molecules simultaneously. By early 2021, when Paxlovid entered Phase I clinical trials, they realized they were quite familiar with this molecule. After reviewing project records, they discovered that they had actually participated in the development of this drug. Ma Jian told his colleagues, "We have taken part in something of historical significance. Though we can participate in the development of countless drugs, it is rare to be involved with one that has such a significant global impact during the COVID-19 pandemic."

A Different Path from Insilico

XtalPi and Insilico Medicine, two AI-driven drug discovery companies that are close to going public, have chosen distinct business models.

AI pharmaceuticals mainly have two business models: AI Biotech companies (AI-Biotech) and AI Contract Research Organizations (AI-CRO). The former uses AI technology to empower pipeline R&D, hoping for efficiency breakthroughs brought by AI pharmaceuticals to establish a first-mover advantage; the latter focuses more on industrial empowerment, with a business logic similar to preclinical CRO companies. In China, the typical example of the former is Insilico Medicine, and the latter is XtalPi.

A senior computational biology researcher working at an AI-Biotech company told the Economic Observer that the ceiling of AI-Biotech is very high, but the floor is also very low, depending entirely on whether the pipeline succeeds. If the target selection is poor or the pipeline encounters unsolvable problems, the R&D investment will be wasted. The revenue of AI-CRO is relatively stable — as long as projects are secured, there will be income. The former’s room for imagination is unknown, while the latter’s is known.

In an interview in April 2024, Ren Feng, co-CEO of Insilico Medicine, said that the business model for AI drug discovery has not yet become very clear, and people are still exploring to see which model can bring real opportunities to AI drug discovery. However, no single business model has been proven to be better than all the others.

Starting from 2023, Insilico Medicine began out-licensing its ongoing biopharmaceutical projects to partners. Ren Feng believes that in terms of generating cash flow, Insilico Medicine aims to advance projects to the preclinical candidate compound (PCC) stage or Phase I clinical trials before transferring them for profit.

The goal of AI pharmaceuticals is ambitious, with practitioners hoping that AI will disrupt traditional paradigms in the pharmaceutical industry. However, this goal has been questioned by many until the advent of ChatGPT (a chatbot developed by OpenAI, a U.S.-based AI company), after which skepticism decreased significantly. A senior executive from an AI pharmaceutical company in China told the Economic Observer that pharmaceuticals involve complex human systems, making it difficult for AI to completely revolutionize the field in a short period. If the disruption is limited to the preclinical stage, AI pharmaceuticals can achieve this in the short term, and XtalPi's model is best suited for participating in preclinical work.

In the winter of the biopharmaceutical industry, many of XtalPi's potential customers in China are disappearing, which is bad news. But at the same time, more and more companies that are "tightening their belts" are beginning to seriously consider using AI to speed up progress and reduce costs. The above-mentioned executive of the AI pharmaceutical company said that if designed well, AI can help cut more than half of the preclinical costs.

"Different Species" from CRO

XtalPi's main businesses are divided into two categories: one is drug discovery solutions, providing solutions across different modules throughout the entire process of drug discovery and research; the other is intelligent automation solutions, mainly including solid-state R&D services and automated chemical synthesis services. XtalPi's clients include 16 of the world’s top 20 biotechnology and pharmaceutical companies, such as Pfizer, Eli Lilly, Johnson & Johnson, and Merck Group. From 2021 to 2023, XtalPi served 75, 120, and 187 clients respectively, with client retention rates of 67.5%, 51.4%, and 64.9%.

As an AI-CRO company, XtalPi is often compared to CRO companies like Wuxi AppTec (603259.SH/02359.HK). Ma Jian said that in fact, XtalPi's genetic makeup and planning are different from those of CRO companies, making it a completely different entity. "Our core competitiveness lies in our ability to integrate and develop new technologies with application scenarios," Ma Jian said. This logic may seem simple, as if one just needs to combine people from field A with technology from field B. However, applying this in real enterprises or projects is very complex, requiring the collaboration and translation efforts of many participants, as well as significant business traction. XtalPi has established unique technological advantages in physical sciences, AI, and automation clusters, possessing professionals and delivery capabilities in many application scenarios, which fosters the occurrence of chemical reactions.

The academic backgrounds of XtalPi's three founders lie in the fields of theoretical physics, quantum computing, and chemistry. Ma Jian said that this gives them a relatively fundamental and technical logic in understanding many issues, while also continuously enhancing their understanding of the industry. They empower drug discovery, materials, and other fields through foundational technologies.

Biopharmaceuticals is a high-risk, high-investment industry. People use the phrase "10 years and 1 billion US dollars" to describe the cost required to develop an original new drug. However, once successfully commercialized, the returns can be very substantial, with many of the top companies in the US being biopharmaceutical firms. How to find a molecule, how to quickly synthesize this molecule in the lab, requires constant trial and error. The faster one experiments, the more likely they are to achieve good results sooner. XtalPi believes it can help pharmaceutical companies run faster in this competitive race.

Ma Jian envisioned two scenarios. One is his dream: designing a molecule in the morning and synthesizing the compound in the afternoon (which currently may take weeks or even months). The other is the goal that XtalPi aims to achieve: humans are still an important labor force in laboratories, and much of scientists' research exploration is not purely intellectual activity but also involves many operational tasks with little scientific value. He believes this part of the work will definitely be largely replaced by more automated and intelligent equipment and platforms.

Recently, XtalPi held a seminar at Merck's Boston R&D center. Many experts and pharmaceutical company executives stopped by the booth, where they could see in real time from XtalPi’s digital management platform the status of every molecule, the stage of the process it was in, and all the information was complete. The seminar took place at 1:00 PM local time in the U.S. and at 1:00 AM in China. Through the live broadcast screen at the venue, people could see that in an unmanned lab in China, much of XtalPi’s work was still ongoing. Ma Jian said this gave everyone a glimpse into the productivity of the future.

|Dialogue|

Economic Observer: How did you and the other two founders decide to start a business together back then?

Ma Jian:We were classmates at the time, often exchanging ideas. I like people who can inspire each other and ignite passion in their work—they give me that sense of connection. When we started our business, we didn’t just design solutions to address specific problems; instead, we ventured into an unknown future. The happiest moment in life is meeting like-minded people in your prime years and working together to accomplish something meaningful.

Economic Observer: How is the division of labor among the three of you?

Ma Jian:Shu Hao is the Chairman, responsible for capital market financing and strategic levels; I am the CEO, in charge of the company's business operations management, R&D, business development, and internationalization; Li Peng is the Chief Innovation Officer, responsible for driving the underlying technology of the company's innovative R&D.

Economic Observer: How to avoid potential divisions or splits that may arise in the future?

Ma Jian:We have many mechanisms to ensure that we reach consensus on important decisions. We classify certain issues into different levels, discuss key points before making decisions, and once a decision is made, it remains unchanged. In our daily work, we rely more on processes and division of labor to support and coordinate with each other.

We complement each other: you do what you like, and I do what I like, forming a well-matched puzzle. Whether in personal life or in managing a business, the issues we encounter often overlap at the boundaries, and the three of us have resolved this very well over the past decade of continuous adjustment. As for new challenges that may arise in the future, I am very confident because we have all become exceptionally mature.

Economic Observer: Have you had any major disagreements?

Ma Jian:The degree of divergence between us is much smaller than the scenario you imagined in your question. At least, I believe we have never had a situation where we were going in completely opposite directions. Within the company, I encourage more "conflict," allowing everyone to communicate and exchange ideas more thoroughly. It doesn't necessarily have to be verbal; written words can also be used to organize one's thoughts.

Economic Observer: On the eve of XtalPi's IPO, some AI pharmaceutical companies in China are laying off staff or even undergoing bankruptcy liquidation. What is the current state of this industry?

Ma Jian:The emergence of a new track will attract a plethora of participants, which precisely demonstrates that this direction is meaningful and valuable. However, market competition and industry development are subject to external cycles and rules, and the growth of a company is influenced by a combination of various factors. It is a normal market phenomenon for some companies to be in the phase of layoffs or even bankruptcy liquidation, and there is no need for over-interpretation. Meanwhile, we do need to look further ahead.

Economic Observer: Peers are all very concerned about commercialization. What are XtalPi's pursuits regarding revenue and profit after its IPO?

Ma Jian:XtalPi has achieved qualified commercialization in many aspects, or has established independent business lines. We need to be relatively patient with new technologies and think long-term, as many of the right things are inherently more difficult to succeed at.

Economic Observer: Recently, Wuxi AppTec has been impacted by developments related to the Biosecure Act in the United States. XtalPi also generates a significant portion of its revenue from its U.S. business. Do you have similar concerns about the future?

Ma Jian:From last year to now, I have traveled to the U.S. five times for business trips. From my personal perspective, our communication and interaction with clients have been very smooth. Although we are sometimes compared to CRO companies, XtalPi is an entirely different entity—our company's DNA and roadmap are distinct. In our international development, we focus more on strengthening our technological capabilities to address the pressing pain points of our clients. Leveraging China’s advantages in manufacturing, industry, and talent, XtalPi has built a unique technological moat. We aim to become the engine and partner for global drug and new material R&D.

Economic Observer: Which companies do you admire? Specifically, in the AI pharmaceuticals field?

Ma Jian:Many companies are XtalPi's benchmarks in certain aspects, but none of them are its ultimate goal. For instance, SpaceX (Elon Musk's space transportation company) pushes engineering technology to the extreme and significantly reduces rocket costs, with the goal of exploring the vast universe. Although XtalPi isn’t heading to Mars, we admire its dream. Another example is OpenAI, which, after its establishment in 2015, accumulated massive amounts of data and invested heavily in model development, finally launching a phenomenal product by the end of 2022. It possesses strong core competitiveness, along with sufficient patience, perseverance, belief, and execution power.

In the AI pharmaceuticals field, we have many excellent peers. However, from the perspective of XtalPi's self-positioning, genetic makeup, and technological foundation, we are actually a technology company. We are committed to addressing numerous engineering and technical challenges, and hope to build the capability to rapidly explore drug molecules through AI + robotics, integrating both hands and brain. This is what we are currently advancing.

Economic Observer: When you founded XtalPi a decade ago, did you ever imagine it would become the first 18C stock and the first AI pharmaceutical stock?

Ma Jian:We have always had big dreams and aimed high. Ten years ago, we had belief in the future rather than confidence, and we knew we had to persevere and push things forward. Today’s listing in Hong Kong—I believe it is either earlier, later, or just right on this day—an inevitable step in the process. The English name of 晶泰科技 is XtalPi, which includes “Pi,” reminiscent of *Life of Pi*. Starting a business is like an adventure with many beautiful sights along the way. Timing, location, and people—all are indispensable elements in the entrepreneurial journey.