−4.8%
since Sep 30
The idea
I believe the economy is starting to pay more for discovery than for knowledge, since AI makes knowing things cheap but makes the rate of new attempts, in drugs, materials, chips, faster than ever. Rather than guess which lab wins, I own the toolmakers who get paid regardless: instrument makers, simulation and AI-for-science software, DNA reading and writing, and lab automation. I spread it across the whole toolkit rather than betting on single names, keep 9% in cash for weakness, and expect to hold this for years, revisiting only if discovery itself slows down.
What it owns
- These companies sell the foundational instruments, consumables and services that labs worldwide depend on for discovery work, collecting revenue across nearly every research attempt regardless of outcome. · Thermo Fisher ScientificIt is the largest global seller of instruments, consumables and services to research labs, so it collects revenue on nearly every discovery attempt across biology and pharma spending., DanaherIt owns a portfolio of life-science and diagnostics businesses that collect revenue as research and drug production scale, with recurring exposure to discovery spending through recent acquisitions., Agilent TechnologiesIt supplies the chromatography, spectroscopy and laboratory instruments that research labs run daily, with sticky recurring revenue from consumables and service contracts on its installed base., Waters CorporationIt dominates liquid chromatography and mass spectrometry with a large installed base and recurring high-margin revenue from consumables, essential to identifying what molecules actually are., BrukerIt sells the heavy scientific instruments including NMR, mass spectrometry and advanced microscopy that frontier research depends on, driven by academic and industrial research budgets., Mettler-ToledoIt makes the precision balances and measurement equipment that sit in nearly every research and quality-control laboratory worldwide, essential for the exact measurements every experiment requires., RevvityIt sells the reagents, instruments and software that life-science laboratories and diagnostics providers consume in daily operations, with recurring revenue from consumables tied to research spending.34%
- These companies sell software that lets researchers test ideas computationally before building and testing them physically, replacing some trial-and-error work with simulation-first design. · SchrodingerIt sells physics-based simulation software that lets drug hunters test molecules on a computer before synthesizing them, enabling computational design to displace wet-lab trial and error., Cadence Design SystemsIt supplies half of the chip-design software duopoly and is extending simulation-first approaches into drug discovery and systems design as design activity grows., SynopsysIt sells the design software that powers modern chip development, and semiconductor research is now central to discovery attempts across materials, quantum and biotech., CertaraIt sells biosimulation software that models how drugs behave in the body, allowing companies and regulators to replace some human and animal testing with computational modeling.19%
- These companies provide the tools and platforms that turn biological designs into physical samples and run the tests that validate them, sitting at the interface between idea and evidence. · IlluminaIt makes the machines that read DNA sequences, providing the raw data layer of biological discovery on a large installed base where revenue scales with sequencing volumes., Twist BioscienceIt synthesizes custom DNA on silicon chips, providing the starting material for biological engineering at lower cost than traditional methods, scaling with demand for synthetic genes., 10x GenomicsIt provides the tools that let researchers study biology at single-cell and spatial resolution, enabling a level of detail that discovery work increasingly demands beyond bulk analysis., Pacific BiosciencesIt reads long stretches of DNA accurately, capturing genetic information that short-read sequencing misses, with revenue tied to adoption of long-read methods in research budgets., Ginkgo BioworksIt operates a platform for engineering organisms, capturing revenue as companies outsource biology R&D attempts to its platform rather than building internal labs.16%
- These companies make the sensors, measurement tools and test equipment that let engineers and scientists observe and quantify phenomena across physical sciences, electronics and industrial applications. · Keysight TechnologiesIt makes the test and measurement equipment that engineers use to develop new electronics, wireless systems and quantum hardware, getting paid on every R&D push forward., Teledyne TechnologiesIt makes the sensors, cameras and measurement instruments that enable scientists and engineers to observe phenomena across industries, from ocean depths to orbit., FortiveIt sells precision instruments and workflow software used across industrial and healthcare operations, capturing revenue as demand for accurate measurement and connected tools expands.13%
- These companies automate the repetitive work of sample management and testing, scaling up the volume of experiments that can run as discovery work accelerates. · AzentaIt stores, moves and manages the biological samples that discovery experiments generate, so sample volumes and automation adoption directly drive its recurring revenue., TeradyneIt tests semiconductor products before shipment and sells collaborative robots for lab and factory automation, capturing revenue as chip production and automation adoption grow.9%
- Cash9%