How the Finch Innovation Index Defines, Tracks, and Scores 73 Investable Technology Themes From Preprint Data
The difference between a useful research intelligence platform and a glorified search engine is methodology. Anyone can count papers. The hard part is deciding what to count, how to group it, and how to turn raw publication volume into a signal that actually informs capital allocation. This post explains how the Finch Innovation Index does each of those things across its 73 investable technology themes.
Defining 73 Investable Themes: Taxonomy Design for Capital Allocators
Most academic taxonomies are built for academics. They follow disciplinary boundaries: physics, chemistry, biology, computer science. That structure is nearly useless for an investor trying to understand whether solid-state batteries or neuromorphic computing is accelerating faster than perovskite solar cells.
The Finch Innovation Index organizes preprint research into 73 themes designed around investable technology categories, not academic departments. Each theme maps to a definable market, a set of venture-backable companies, or a strategic national priority. Examples span AI, biotech, climate tech, quantum computing, advanced materials, robotics, energy storage, and dozens of other verticals.
The Finch Innovation Index covers 73 investable technology themes spanning AI, biotech, climate tech, quantum, advanced materials, and other verticals. Theme boundaries are defined using curated keyword sets, classifier models trained on labeled preprint abstracts, and periodic manual review by domain specialists. This is not a static dictionary lookup. Themes evolve as research language shifts, which is why the system also tracks rising keywords and emerging clusters that may eventually warrant their own theme designation.
The taxonomy is designed so themes are mutually informative but not rigidly exclusive. A preprint on protein structure prediction may appear under both AI and biotech themes, because it carries signal for both. For a deeper look at how research maturity varies across these themes, see Technology Readiness Across 73 Investable Themes.
Tracking: From Raw Preprints to Structured Monthly Signals
The Finch Innovation Index processes over 1 million classified preprints to generate monthly intelligence. Sources include arXiv, bioRxiv, medRxiv, ChemRxiv, and other major preprint servers. Each paper is ingested, classified into one or more of the 73 themes, and tagged with metadata including author affiliations, geographic origin, and submission date.
Monthly snapshots capture several dimensions per theme: absolute volume (how many preprints were published), growth rate (how volume is changing relative to prior periods), geographic distribution (which countries and institutions are producing work), and keyword velocity (which terms are rising or declining within the theme's corpus).
This structured tracking is what separates preprint intelligence from a literature review. The output is not a reading list; it is a quantitative time series that can be benchmarked, compared, and modeled. Investors who want to understand how this maps to commercial potential should read From arXiv to Investment Thesis.
Scoring: What Momentum Captures and Why It Matters
The core analytical output of the Finch Innovation Index is the momentum score. Momentum scoring in the Finch Innovation Index measures the rate of change in research output, not just the absolute level. A theme with 500 preprints per month that has been flat for two years scores lower than a theme with 80 preprints per month that has doubled in six months.
This distinction matters enormously for investment timing. Absolute volume tells you where the established research communities are. Momentum tells you where new ones are forming. The Finch dataset provides a 2 to 5 year signal advantage over traditional patent or market-based indicators, precisely because preprints appear years before patents are filed or products reach market.
Momentum scores are normalized to allow cross-theme comparison. You can directly compare whether quantum error correction is accelerating faster than gene therapy delivery, even though the two themes differ by an order of magnitude in publication volume. For a full explanation of the scoring mechanics, see how momentum scoring works in research intelligence.
Geographic Intelligence: Who Is Publishing, and Where
Beyond theme-level momentum, the Finch Innovation Index surfaces geographic concentration patterns for each theme. Geographic signals in the Finch Innovation Index reveal which countries dominate specific technology themes at the research stage. For some themes, research is globally distributed. For others, two or three countries account for the vast majority of output. These patterns matter because research concentration at the preprint stage tends to predict commercial concentration five to ten years later.
Each theme's geographic profile is updated monthly, allowing analysts to track shifts in national research priority. A sovereign wealth fund evaluating semiconductor supply chain risk, for example, can observe whether preprint output in advanced lithography is diversifying or consolidating geographically, years before the commercial implications are visible.
Why Methodology Transparency Matters for Users
The Finch Innovation Index publishes its methodology because the value of any index depends on whether users trust its construction. Black-box scores invite skepticism. Transparent taxonomies, classification logic, and scoring formulas invite scrutiny, which is exactly what sophisticated institutional investors expect before integrating a signal into their workflow.
The Finch Innovation Index is built for analysts who understand that research output is the earliest measurable signal of technological change. The methodology described here is the mechanism that turns that raw signal into structured, actionable intelligence across 73 themes, every month.