Analysis of Insilico Medicine's Dark Pool Surge, AI Pharma Valuation, and Hong Kong IPO Performance

Based on the information I collected, I provide a detailed analysis of the reasons behind Insilico Medicine’s dark pool surge, the valuation logic of AI pharmaceutical companies, and the sustainability of Hong Kong IPO performance.

The artificial intelligence sector continued to heat up in 2025, and AI pharma, as one of the most promising tracks in the healthcare field to benefit from AI, attracted significant capital attention [1]. On December 23, another AI company Nobikhan closed up 363.75% on its first day of listing, setting a new high for the largest first-day gain of new stocks this year, reflecting the secondary market’s sustained enthusiasm for tech stocks, especially AI concepts [3].

Hong Kong’s IPO market saw a strong recovery in 2025. Deloitte statistics show that as of December 17, there were 114 new listings in Hong Kong, raising approximately HK$286.3 billion, more than three times that of 2024. Among them, the healthcare and pharmaceutical industry raised HK$25.2 billion, an increase of HK$22 billion compared to 2024. Hong Kong even returned to the top spot globally in IPO fundraising in 2025 [1].

The Hong Kong Securities and Futures Commission (SFC) and the Hong Kong Exchanges and Clearing Limited (HKEX) continued to optimize policies around key areas such as listing thresholds, approval efficiency, and market liquidity, boosting Hong Kong’s IPO into a new cycle. As of December 17, 2025, 19 A+H new stocks were listed, accounting for about 50% of the total fundraising [3].

Insilico Medicine’s self-developed generative AI platform Pharma.AI is its core competitive advantage. The traditional drug discovery process usually takes 3 to 6 years, but from 2021 to 2024, Insilico Medicine has nominated 20 preclinical candidate compounds, with an average of only 12 to 18 months from project initiation to preclinical candidate nomination, and only 60 to 200 molecules synthesized and tested per project, significantly improving R&D speed and resource utilization efficiency [9].

The company’s fastest-progressing candidate drug Rentosertib (ISM001-055) for the treatment of idiopathic pulmonary fibrosis has entered Phase IIa clinical trials and is the world’s fastest-progressing AI drug [7]. The target TNIK of this drug was discovered with the assistance of artificial intelligence, which has attracted much attention from the industry [7].

Insilico Medicine has generated more than 20 clinical or IND application-stage assets through the Pharma.AI platform, three of which have been licensed to international pharmaceutical and healthcare companies, with a total contract value of up to $2.1 billion, including approximately $110 million in upfront payments and about $1.9 billion in milestone payments [7]. In November 2025, the company entered into a strategic cooperation with Eli Lilly, entitled to receive up to $100 million in revenue from this cooperation [9].

According to the prospectus data, Insilico Medicine’s revenue from 2022 to 2024 was $30.147 million, $51.18 million, and $85.834 million, showing a rapid growth trend; the net losses during the same period were $220 million, $210 million, and $17.1 million, respectively, with a significant narrowing of losses [7].

Insilico Medicine’s global offering of 94,690,500 shares, with 10% for Hong Kong public offering and 90% for international offering, at an issue price of HK$24.05 per share, with an IPO market capitalization of approximately HK$13.406 billion, making it the largest Biotech IPO in Hong Kong’s market in 2025 [5].

Hong Kong’s new stock subscription enthusiasm was high in 2025, with 38 new stocks subscribed more than 1000 times during the year, among which Gold Leaf International Group listed in October even exceeded 10,000 times. The average public subscription multiple jumped from about 150 times in the first half of the year to 3175 times after August [3].

As the “First AI Pharma Stock”, Insilico Medicine has scarcity in the Hong Kong market, and investors have given a scarcity premium.

The traditional DCF (Discounted Cash Flow) valuation method faces huge challenges in the AI pharmaceutical field:

1. Extremely limited predictable cash flow, as most companies are in preclinical or early clinical stages
2. Exponentially increasing uncertainty in technology and commercialization
3. The discount rate curve rises steeply in the terminal value stage, easily leading to valuation bubbles
4. The NPV indicator itself has no time dimension; it only answers “whether it is worth it” but not “how long to wait” [6]

Most AI pharmaceutical companies have not yet achieved profitability, and even their revenue scales are small, making traditional P/E and P/S indicators difficult to effectively evaluate. For Insilico Medicine, its 2024 revenue was about $85.83 million, but its market capitalization was still as high as about $1.7 billion (calculated at the issue price), with a P/S ratio of about 20 times [2][5].

• Algorithm innovation: Whether it has core algorithms with independent intellectual property rights
• Data advantages: Whether it has unique high-quality biomedical data sets
• Integration capability: Whether it can integrate multi-disciplinary data such as biology, chemistry, and clinical medicine
• Iteration speed: Frequency and effect of model optimization and algorithm updates

Insilico Medicine’s Pharma.AI platform integrates modern machine learning technologies such as deep generative models, reinforcement learning, and transformation models to build an integrated drug R&D platform [9].

• Time cycle from target discovery to preclinical candidate drug (PCC) (Insilico Medicine: 12-18 months vs traditional: 4.5 years)
• Project success rate (candidate compound screening efficiency)
• Number and quality of patents (Insilico Medicine holds a total of 787 patents and patent applications globally) [5]

Insilico Medicine has established more than 30 independent R&D pipelines, covering multiple fields such as fibrosis, oncology, immunology, metabolic diseases, and neurological diseases [7]. Diversified layout reduces the risk of failure of a single project.

The core asset Rentosertib has entered Phase IIa clinical trials, which is a milestone of great significance as the world’s fastest-progressing AI drug [7]. The later the clinical stage of the pipeline, the higher the valuation weight.

The total value of external licensing cooperation agreements (Insilico Medicine: up to $2.1 billion), the receipt of upfront payments, and the trigger conditions for milestone payments are all important valuation references [5][7].

Insilico Medicine’s revenue mainly comes from three business segments:

1. Drug discovery and pipeline development (accounting for 92.9% in 2024): Core revenue source
2. Software solutions (accounting for 4.6% in 2024)
3. Other discovery businesses (accounting for 2.5% in 2024): Non-pharmaceutical applications [5]

From 2022 to 2024, Insilico Medicine’s revenue from the top five customers accounted for 90.6%, 94.1%, and 94.4% respectively, with a certain degree of large customer dependence risk [7].

Whether the team has compound talent backgrounds in drug R&D, AI algorithms, clinical development, etc.

The cash reserve situation of Insilico Medicine as of June 30, 2025 (detailed information needs to be checked in the prospectus) is crucial to support continuous R&D investment.

• Interpretability of AI models and regulatory acceptance
• Data quality and bias issues
• Algorithm substitutability (competition from open-source models)

• Clinical trial success rate (traditional drugs have a success rate of about 10-15% from Phase I to launch)
• Long R&D cycle (even with AI support, it still takes 5-10 years)
• Huge R&D investment

• No AI-developed drug has been successfully launched globally
• Acceptance and willingness to pay for AI technology by pharmaceutical companies
• Competition with traditional CRO/CDMO

• Valuation bubble risk (some AI concept stocks have market capitalization hundreds of times their revenue [1])
• Investor patience and capital sustainability issues
• Intensified industry competition

It is recommended to adopt a multi-dimensional comprehensive assessment method:

Assessment Dimension	Key Indicators	Recommended Weight
Technology Platform	Algorithm innovation, data advantages, R&D efficiency improvement	30%
Pipeline Value	Number of clinical-stage pipelines, core asset progress, cooperation value	30%
Commercialization	Revenue growth, customer quality, gross margin trend	20%
Team Resources	Team background, cash reserve, patent layout	15%
Risk Factors	Technology risk, R&D risk, market competition	5%

Dark pool trading refers to over-the-counter trading conducted through the internal system of brokers after the market closes (usually 16:15-18:30) on the trading day before the new stock listing, not through the HKEX trading system [4].

1. Reflects the expected pricing of new stocks by institutional investors and retail investors on the first day
2. Provides opportunities for early profit-taking/stop-loss
3. Reflects the market enthusiasm for new stocks to a certain extent

1. Only reflects the trading situation of customers of specific brokers, not the entire market
2. Relatively limited liquidity
3. Prices may be affected by individual large investors

According to Futu Securities data, screening Hong Kong IPOs listed from December 1, 2024 to May 23, 2025, the probability of closing with a gain of ≥0% on the first day is 79.4%, and nearly 80% of Hong Kong IPOs rose on the first day of listing [4].

• Nobikhan (02635.HK): Dark pool surged, closed up 363.75% on the first day of listing
• Easy Health (02661.HK): Dark pool surged, closed up 158.82% on the first day
• Gold Leaf International Group (08549.HK): Dark pool surged, closed up 330.0% on the first day [2][3]

However, some new stocks have diverged between dark pool performance and first-day performance:

• BenQ Hospital (02581.HK): Fell 49.46% on the first day
• Huazi Bio (02396.HK): Fell 29.32% on the first day
• Hansi Aite-B (03378.HK): Fell 46.25% on the first day [2]

Statistics show that public subscription enthusiasm is mostly positively correlated with the first-day performance of new stocks—the more “popular” the new stock, the better the first-day performance. After August 2025, the average public subscription multiple of new stocks rose rapidly from about 150 times in the first half of the year to 3175 times [3].

Endorsement by well-known institutions such as Tencent and Sequoia can enhance credibility. The cornerstone investor lineup of Insilico Medicine (detailed information needs to be checked in the prospectus) has an important impact on market confidence.

New stocks with “routine clawback” (under-subscribed international offering but over-subscribed public offering) have relatively better first-day performance. Since 2024, 22 new stocks have been under-subscribed in international offerings but over-subscribed in public offerings, with clawback ratios between 10% and 20%, among which 20 rose or closed flat on the first day of listing, with a break rate of only 9.09% [4].

Hong Kong’s market has recovered overall in 2025, and technology stocks and biomedicine sectors are favored by capital, which is conducive to the sustainability of new stocks’ performance after listing.

• New stocks that surge in dark pool usually perform well on the first day, but may diverge after the first day
• Some new stocks pull back after surging on the first day; it is recommended to pay attention to trading volume and turnover rate

• Fundamentals (business progress, financial data) become the dominant factor
• Industry prosperity and market style transformation have a greater impact

• Among the most profitable new stocks in 2025, Yaojie Ankang-B (02617.HK) soared for several consecutive days after being included in the Hong Kong Stock Connect, once rising more than 50 times compared to the issue price at the highest point, and still rose about 890% compared to the issue price as of the close on December 24, ranking first in the annual new stock gain [3]

• Statistics show that 85% of Hong Kong new stocks break within one year; it is recommended to close positions on the first day or use the strategy of分批 selling in dark pool/first day to smooth returns [4]

1. Sustained enthusiasm for AI concept, broad track prospects
2. Fast business progress, core asset in Phase II clinical trial
3. Cooperation with large pharmaceutical companies such as Eli Lilly, verifying technical strength
4. Significant growth in Hong Kong Biotech IPO fundraising in 2025, improved market recognition

1. Has not yet achieved commercial profitability, relying on continuous financing
2. No AI drug has been successfully launched globally, with technical verification risks
3. Revenue from the top five customers accounts for more than 94%, with concentration risks
4. The issue market capitalization of HK$13.4 billion needs performance support

• Short term: Pay attention to the performance on the first day and first week; the dark pool surge provides an opportunity for early profit-taking
• Medium term: Pay attention to core drug Phase II clinical data and cooperation progress
• Long term: Wait for the commercial verification of AI drugs and pay attention to the overall progress of the industry

Insilico Medicine’s 201% dark pool surge reflects the market’s high expectations for the AI pharmaceutical track. This sentiment-driven rise has its rationality (technological progress, business progress, market environment) but also has a certain degree of speculation.

For investment in AI drug R&D companies, it is recommended:

1. Rationally view technology hype and focus on substantive progress
2. Adopt a multi-dimensional valuation framework instead of simply applying traditional valuation methods
3. Recognize that this is a high-risk and high-return track that requires a long-term perspective
4. Pay attention to key milestones such as clinical data verification and commercial progress

For Hong Kong new stock subscription, it is recommended:

1. Dark pool performance can be used as a reference but not the only standard
2. Pay attention to indicators such as public offering subscription multiple, cornerstone investors, and clawback mechanism
3. Short-term profits can be considered for分批 selling in dark pool or first day
4. Long-term holding requires in-depth research on company fundamentals

Whether Insilico Medicine can continue to rise ultimately depends on whether its AI drugs can successfully pass clinical trials and achieve commercialization. As an investor, one should not only see the huge potential of AI pharmaceuticals but also fully recognize the challenges and risks it faces.

[0] Jinling API Data

[1] Yahoo Hong Kong Finance - “Chinese AI Concept Attracts Capital Hype but Behind the Excitement are Technical Bottlenecks to be Broken Through” (https://hk.finance.yahoo.com/news/)

[2] East Money - “Hong Kong IPO Weekly Report: Companies such as ViewSonic Co., Ltd. Submit Applications Collectively, Multiple New Stocks Break on First Day” (http://finance.eastmoney.com/a/202512283603755364.html)

[3]华盛通 - “Year-End Inventory | The Most Profitable New Stock List is Released! The Highest Profit per Hand is 14,600 Hong Kong Dollars, Surging Over 360% on the First Day, How Fierce is Hong Kong New Stock Subscription in 2025?” (https://www.hstong.com/news/detail/25122423382818737)

[4] Futu Niuniu - “Subscribe to New Stocks with Futu, No.1 New Stock Subscription Broker” (https://invest.futuhk.com/hk_ipo)

[5] East Money - “The Largest Biotech IPO in Hong Kong This Year! Insilico Medicine Plans to Raise HK$2.3 Billion to Sprint for the First AI Pharma Stock” (https://finance.eastmoney.com/a/202512183595647279.html)

[6] Shifeng Research - “Yang Chuan: Valuation Revolution - From Free Cash Flow to Growth-Driven Value Model (9)” (https://www.shifd.net/yanjiu/detail/10229.html)

[7] Securities Times - “Competition in AI Pharma Track Intensifies, Insilico Medicine Breaks Through in Its Third Attempt at Hong Kong IPO” (https://stcn.com/article/detail/3381348.html)

[8] Investment Notes Network - “AI Pharma Under the Wave of Large Models: Startups Secure $1 Billion in Financing” (https://www.phirda.com/artilce_39225.html)

[9] EurekAlert! - “Insilico Medicine Partners with Eli Lilly to Reach Strategic Cooperation in AI-Driven Drug R&D” (https://www.eurekalert.org/news-releases/1105380)

[10] Gelonghui - “A Decade-Long Opportunity for Hong Kong New Stock Subscription! One Article Explains Everything You Need to Know” (https://www.gelonghui.com/news/5022334)