J.P. Morgan: Maintains "Overweight" rating on INSILICO (03696), target price HK$71

J.P. Morgan released a research report stating that INSILICO's (03696) recent progress in longevity research reinforces its differentiated advantage in the AI drug discovery (AIDD) industry; the bank maintains its "Overweight" rating on INSILICO with a target price of HK$71, based on a sum-of-the-parts (SOTP) valuation as of December 2027, in which drug discovery and software solutions and other businesses are assigned 17.5x and 8.9x EV/S multiples respectively, implying a 2027 forecast P/S ratio of approximately 19x, at the lower end of the global AIDD peer range of 5x to 173x (median approximately 60x). The bank noted that last Thursday (the 17th), a Cell cover study introduced the aging-related benchmark dataset LongevityBench, specialized longevity language models Longevity-LLMs, and the Longevity Claw research agent, establishing an integrated set of tools for evaluating AI reasoning and identifying potential aging-related targets. This complements a recent Nature Biotechnology publication on interesting aging biomarker changes observed in patients with idiopathic pulmonary fibrosis (IPF) treated with INSILICO's Phase 3 asset rentosertib. The bank believes these developments link computational discovery with emerging human evidence. The bank noted that INSILICO's share price has risen 76% since its mid-June low (during which the Hang Seng Healthcare Index (HSHCI) rose 24%), reflecting strong market sentiment toward its AIDD capabilities and commercialization potential. The bank believes its investment thesis remains attractive, as the company combines clinical assets approaching key validation, demonstrated out-licensing potential, and additional opportunities in aging-related diseases. The bank continued that the aforementioned Cell paper reinforces INSILICO's approach to discovering drugs targeting aging biology. The study introduced three complementary tools: LongevityBench assesses AI models' ability to interpret aging-related biological data; specialized language models analyze clinical and multi-omics datasets; and Longevity Claw combines these models with scientific tools to identify and prioritize therapeutic targets. The specialized models matched or surpassed the general-purpose large language models tested on the research benchmark, highlighting the importance of domain-specific data and training. Longevity Claw nominated 328 potential intervention targets, with enrichment up to 5.6x relative to an independent, experimentally supported reference set of aging targets. The bank believes this is a more systematic way of translating complex aging data into testable drug discovery hypotheses; experimental validation remains necessary, but this capability can broaden INSILICO's pipeline and create additional collaboration opportunities. Rentosertib remains INSILCO's most clinically advanced example targeting the shared mechanisms of aging and specific diseases. In another Nature Biotechnology study, an exploratory analysis of 42 patients from its Phase 2a IPF trial (n=128) found that six proteomic aging clocks showed lower predicted biological age in the treatment group. The bank believes the consistency of results across models reinforces the rationale for further research, although significance varied by dose and time point, and the analysis cannot fully distinguish effects on aging from lung disease improvement. The TNIK inhibitor rentosertib has now entered Phase 3 in China, with first patient dosing announced on September 10. The bank views this as a commercially pragmatic development strategy: establishing efficacy in a well-defined disease while studying broader biological effects. Success in IPF development would reinforce the rationale for exploring additional indications, but longevity applications require their own clinical evidence.