In the intensifying wave of AI-driven drug discovery, the ultimate biggest beneficiary may well be CROs.

Recently, the hottest concept in the biopharmaceutical sector is undoubtedly AI pharma. At the end of last month, the latest international "story" emerged: a former Goldman Sachs analyst built a one-person pharma company called Mithril Medicine, and Anthropic stepped in to acquire it. Because this single-person pharma company does not build its own laboratory, CROs (Contract Research Organizations, the core track of CXO, undertaking outsourced R&D and clinical trial operations for pharmaceutical companies) have instead become the center of gravity in the overall new drug R&D process. Mithril Medicine itself also acknowledged that the current pace of project advancement is not constrained by AI molecular design capability, but is highly constrained by the experimental delivery efficiency of CROs and the offline decision-making speed of scientific personnel.

Such examples are not isolated, which may indicate that in the process of AI reconstructing the organizational paradigm of pharmaceutical companies, CROs may become the biggest winners.

Meanwhile, narrowing down to the listed company level, as the 2026 third-quarter earnings disclosure period concluded, targets in this sector have reported good news one after another. For example, leading company WuXi AppTec reported in its Q3 results that revenue for the first three quarters was 45.537 billion yuan, up 38.59% year-on-year; net profit attributable to shareholders was 19.398 billion yuan, up 60.64% year-on-year. Third-quarter single-quarter revenue was 16.640 billion yuan, a record high since listing; single-quarter net profit attributable to shareholders was 8.318 billion yuan, up 136.67% year-on-year. But doubling net profit is not the peak within the sector. JOINN (ASX: 06127), which focuses on life sciences, reported revenue of 1.188 billion yuan for the first three quarters of 2026, up 20.60% year-on-year; net profit attributable to shareholders of 930 million yuan, up 1,051.94% year-on-year; and non-GAAP net profit of 881 million yuan, up 2,933.08% year-on-year.

In a written response, Tan Xiongtian, a member of the investment committee at Puzhuo Capital, stated: "Judging from this year's third-quarter reports of listed companies, the CRO sector as a whole shows characteristics of differentiated recovery: leading enterprises have stabilized orders and capacity utilization is slowly recovering; small and medium-sized CRO companies show greater earnings elasticity, but some companies in certain sub-sectors still face issues such as delayed overseas client orders and differentiated gross margin recovery pace. The recovery rhythm of preclinical CROs is overall faster than that of late-stage clinical CROs."

AI Pharma Surges as CRO Sector Welcomes New Performance Opportunities

The most sensational event recently in the international pharmaceutical industry circle is AI computing giant Anthropic's pre-IPO acquisition of a biopharmaceutical startup called Mithril Medicine. According to investor disclosures, this company is called Mithril Medicine and was founded only 18 months ago, focusing on the pediatric rare disease LAMA2 congenital muscular dystrophy, with the goal of developing an antibody. The founder previously worked as a Goldman Sachs analyst and studied biotechnology and an MBA at Harvard. The company's founder used Claude to coordinate 11 external laboratories, handling experimental data, project scheduling, and daily operations, single-handedly keeping the company running. From proposing a therapeutic concept to completing preliminary animal experiment validation took less than 18 months. Specifically, he had Claude manage contracts and logistics, distributed experiments to 11 external laboratories, and invited 8 part-time consultants to advance together. Mouse experiments have already shown results, and Anthropic plans to enter human trials before the end of 2028.

And such collaborations are not limited to this one company. Anthropic officially confirmed that over the past few months it has launched cooperation with 6 new drug R&D institutions. Through this newly popular model, it is not difficult to see that in the intensifying wave of AI pharma, the ultimate biggest beneficiary may be CROs. Gao Chao, a special researcher at Guangzhou Minggui Capital, analyzed that the most valuable thing in this model is that outsourcing list. Antibody production, binding tests, cell experiments, animal experiments — all split among 11 external laboratories. Within one week, 5 CROs took on 113 molecular constructs, performing antibody function validation, druggability assessment, and mouse pharmacokinetics. In a 38-page report, CRO appears 58 times, while AI/Claude appears 47 times. An AI company preparing for IPO wrote about outsourcing more than about AI in its own technical background document. This ratio is more intuitive than any brokerage estimate. AI output must be converted into wet-lab validation to count; between "computing" and "doing," money flows toward the "doing" side.

In this regard, Tan Xiongtian analyzed that this marks large model companies deeply entering AI drug discovery, with the business model landing on "AI molecular design + outsourced CRO experimental execution" rather than building their own laboratories. This industry signal is more of an expectation-level catalyst for A-share CROs, AI pharma, and experimental animal sectors. The market will reprice the value of early-stage drug discovery CROs; compared with back-end clinical CROs, front-end services such as molecular synthesis and in vitro pharmacology will receive higher attention. The logic for AI pharma targets is shifting from simply "large models telling stories" to the experimental execution capability after validating AI-output molecules. Targets with only algorithms and no mature CRO integration capability will continue to face valuation pressure.

At the domestic industry level, recent highlights in this sector are reflected in accelerated industrial M&A integration and the explosion of the AI pharma dry-wet closed loop. Regarding the former, several landmark events have recently appeared in the industry, with three involving listed companies: First, Shanghai main board listed company Baihua Pharmaceutical transferred controlling stake for 890 million yuan to Jinhua state-owned assets, representing continued positioning by local state-owned capital in CRO platforms; second, ChiNext listed company Wanbang Pharmaceutical acquired 75.52% equity of Saidesheng for 302 million yuan, a typical mid-sized CRO horizontal completion of the clinical segment; additionally, Hong Kong-listed WuXi XDC acquired Dongyao Pharmaceutical, successfully securing mature ADC capacity and commercialization qualifications, a capacity-positioning M&A in the popular conjugate drug sector.

Regarding the latter, AI does not reduce experiments; it increases the number of molecules awaiting validation. Estimates suggest that in the next 3 to 5 years, the number of projects entering wet-lab experiments will increase by about 12%, meaning that AI has not replaced experiments but pushed more molecules to the stage where experiments must be done.

Order-Driven Sector as Listed Companies Deliver Strong Q3 Results

Moving from the industry level back to the secondary market level, listed companies in the CRO sector together with innovative drugs have become two bright landscapes within the sector: On the semi-annual line, Wind's CRO index has recorded five consecutive positive candles, rising about 3.7% since the second half of this year. Analyzing the reasons for the sector's continued strength, apart from the aforementioned industry-level breakthroughs, the biggest fundamental surprise is the steady upward trend in performance.

From the recently concluded 2026 third-quarter reports of listed companies, according to a Huaan Securities research report, 41 companies reported revenue of approximately 103.3 billion yuan for the first three quarters, up 25.5% year-on-year; net profit attributable to shareholders of approximately 25.82 billion yuan, up 39.5% year-on-year; and non-GAAP net profit up 58.6% year-on-year. Breaking down by quarter, single-quarter revenue all increased year-on-year: Q1 up 18.2%, Q2 up 28.6%, Q3 up 28.8%. Similarly, quarterly net profit attributable to shareholders also all increased year-on-year: Q1 up 24.1%, Q2 up 12.7%, Q3 up 87.9%.

"In this round of CRO third-quarter reports, gross margin is the most outstanding data point for most leading and sub-sector leader companies. The core reasons behind this can be divided into three layers: First, during the industry trough over the past two years, many companies suspended capacity expansion and shut down inefficient laboratories, releasing fixed cost pressures such as depreciation and labor; after capacity utilization recovered, unit fixed costs were diluted. Second, the bargaining environment with overseas clients improved marginally, low-price orders from earlier intense price wars gradually expired, and pricing for new orders recovered. Third, represented by preclinical CROs like JOINN, the business structure continued to shift toward high-gross-margin pharmacology and toxicology safety assessment businesses, while the proportion of low-gross-margin basic contract manufacturing business declined. Combined with new front-end molecular experiment orders brought by AI pharma companies like Mithril, which carry higher technical value-added, this further boosted overall gross margin levels, becoming the most impressive improvement item on the financial statements," Tan Xiongtian commented on the financial reports.

Looking further at several data extremes among targets: From the third-quarter reports, the highest total revenue for the first three quarters and the highest single-quarter Q3 revenue both belong to WuXi AppTec, at 45.537 billion yuan and 16.640 billion yuan respectively. Moreover, the company's backlog for continuing operations reached 76.74 billion yuan, also the highest among sector companies. As for the highest year-on-year growth rate in net profit attributable to shareholders and the highest year-on-year growth rate in non-GAAP net profit, the former belongs to JOINN, and the latter belongs to Innostellar.

In a written response, Gao Chao pointed out that orders are the most critical item: "Innostellar signed new orders of 2.218 billion yuan, up 183.27% year-on-year; JOINN signed 3.42 billion yuan in new orders for the first three quarters, up 109% year-on-year, with backlog of approximately 4.5 billion yuan, up 80% year-on-year; Medicilon signed 1.233 billion yuan in new orders in the first half, up 104.74% year-on-year; Pharmaron's new orders increased by more than 30% year-on-year; WuXi AppTec's backlog increased by 36.4%, with a sequential increase of approximately 10 billion yuan."

From the perspective of institutional positioning, according to brokerage research reports summarizing public fund quarterly reports, the number of funds holding heavily weighted pharmaceutical stocks increased from 1,115 at the end of Q2 to 1,727, with the pharmaceutical allocation ratio approaching the end of Q3 2025 (9.7%). The top three in terms of market value increase were CXO companies, with Asymchem and JOINN also in the top ten. The direction of market value changes indicates allocation tilting toward CXO.

Combined with institutional research, listed companies in the sector are also the focus of sell-side attention led by public funds, such as Pharmaron. Wind data shows that the company held two video conference research sessions on August 21 and October 28, with a total of 53 institutions participating, including 21 fund companies, the largest number among sell-side participants. From the most recent research record, institutions mainly asked layered questions around the company's performance. For example, when answering "What are the main drivers of the notable acceleration in laboratory services segment orders in Q3 2026?", Pharmaron stated: "The company's laboratory services business, already at a globally leading scale in the industry, has consistently maintained growth above the industry average. Order drivers mainly include: First, capturing new opportunities on the foundation of traditional small molecule advantages, including complex small molecules, PROTACs, molecular glues, as well as rapidly developing areas such as complex large molecule conjugates, cyclic peptides, and small nucleic acids; Second, continued progress with strategic clients, with revenue from MNC clients in the United States, Europe, and Asia-Pacific continuing to increase; Third, diversified client structure."

When answering "What is the growth trend of new orders in the company's small molecule CDMO services segment? From late-stage projects and new order reserves, which sub-sectors will contribute more significantly to future performance?", Pharmaron stated: "In terms of new orders, in the first three quarters, the company's CDMO segment new orders grew by more than 40% year-on-year, continuing a strong growth trend. In Q3 2026, the company's API production facility in Ningbo once again passed FDA on-site inspection, providing strong assurance for the company's continued service to U.S. and global markets for commercial innovative drug production. The company's small molecule CDMO business overall has growth continuity. The company's CRDMO pipeline serves nearly a thousand R&D projects annually. With the continued advancement of the small molecule CDMO pipeline, we expect demand for mid-to-late clinical and commercial production projects to continue maintaining strong growth."

Controllable Raw Material Costs Drive JOINN's Outstanding Profit Performance

Descending from the sector level to the company level, apart from long-term stable performer WuXi AppTec, some sub-sector leaders have become dark horses in the CRO supply chain, such as domestic preclinical safety assessment leader JOINN. Wind data shows that after doubling for the full year last year, its stock price continued to strengthen in the secondary market this year: up about 9.17% in the first half, and up about 4% since the second half.

Analyzing the reasons for the continued stock price strength, the primary factor is still performance. From the company's third-quarter report, in the first three quarters, JOINN achieved revenue of 1.188 billion yuan, up 20.60% year-on-year; net profit attributable to shareholders of 930 million yuan, up 1,051.94% year-on-year; non-GAAP net profit attributable to shareholders of 881 million yuan, up 2,933.08% year-on-year; and net cash flow from operating activities of 853 million yuan, up 220.31% year-on-year. For Q3 alone, JOINN achieved revenue of 484 million yuan, up 53.01% year-on-year; net profit attributable to shareholders of 182 million yuan, up 821.27% year-on-year; and non-GAAP net profit attributable to shareholders of 182 million yuan, up 2,938.46% year-on-year.

In this regard, Tan Xiongtian pointed out: "Apart from WuXi AppTec, JOINN is a highly representative target for this quarter. JOINN's core business focuses on non-clinical drug evaluation, namely preclinical safety assessment, which aligns with the most upstream experimental validation segment of the AI pharma supply chain. In the third-quarter report, we can see recovery in overseas client orders, continued volume growth in domestic innovative drug projects, and the gradual release of the company's capacity layout in non-human primate experimental animals, with revenue growth outperforming many pure back-end clinical CRO companies in the sector. At the same time, the company continued cost control, optimizing expenses in animal procurement and laboratory operations. Compared with some small companies in the industry still in losses with declining revenue, JOINN, relying on the rigid demand attribute of safety assessment, was among the first to realize bottom-of-cycle recovery in performance, becoming a preclinical CRO target that benefits greatly from the AI new drug wave."

However, Gao Chao is relatively cautious: "JOINN's profit growth rate is more than 50 times its revenue growth rate. A company doing safety assessment is unlikely to make profit 50 times higher by taking more orders. Of the 930 million yuan in net profit attributable to shareholders, 866 million yuan is fair value changes of experimental monkeys. This could be interpreted as the strongest cyclical elasticity, or as profit relying on monkeys — it should be viewed objectively."

Specifically, the company is a "monkey resource barrier" CRO bound to core production factors. It has over 50,000 cynomolgus monkeys in its own colony, 90% self-bred, making it the safety assessment enterprise with the largest experimental monkey reserve in China. Therefore, against the backdrop of continuous experimental monkey price increases, it completely controls the most core rigid production factor of the safety assessment business, thoroughly avoiding the cost impact of soaring monkey prices in the industry, and building cost and delivery barriers that other CROs cannot replicate in the short term.

Returning to the CRO supply chain level, Tan Xiongtian analyzed: "Currently, the CRO sector still faces several core problems: First, uncertainty in overseas orders caused by fluctuations in overseas pharmaceutical companies' capital expenditure and geopolitical policies; second, homogenized competition among a large number of small and medium-sized CROs, with low-price involution making gross margins prone to fluctuation; third, in the face of new AI pharma demand, many CROs lack standardized experimental processes suitable for AI molecular high-throughput screening, making it difficult to undertake AI-driven new drug projects like Mithril."

He also offered personal solutions: "First, diversify the business structure, continue deepening domestic innovative drug clients, and reduce the risk of single overseas client or single region orders; second, abandon low-price homogenized competition and shift toward high-value-added sub-sectors such as preclinical safety assessment, ADC, and rare disease drugs, establishing differentiated technical barriers; third, proactively transform laboratory processes, build high-throughput experimental systems, open integration interfaces with AI pharma companies, undertake large-batch validation experiments for AI-generated molecules, seize new orders brought by the AI new drug industry explosion, and complete the upgrading of business capabilities."

(This article was published in the November 7 issue of Securities Market Weekly. Stocks mentioned are for illustrative analysis only and do not constitute investment advice.)

For more in-depth original content, click the links below: Market Value Management Series | ESG Series Reports. Finance | Consumer | Traditional Industries | Emerging Industries | IPO