See what themes move markets.
Onthos' Thematic Factor Models (TFM) reveal the thematic drivers behind stock returns, enabling precise return attribution and risk management beyond traditional models. Available in two complementary forms: a fundamental model (TFM-F) and a statistical model (TFM-S).
TFM decomposes market movements into interpretable thematic drivers, giving a view of what moves stock prices that traditional factor models do not provide.
Thematic factors in two complementary forms. Both are daily, global and point-in-time; they differ in how the factors are built and what they are best used for.
Estimates a return for each of 200+ themes directly from stock returns, using companies' business exposures as factor loadings. Powered by our proprietary Taiji estimator and Time Zone Overlap (TZOL) framework for robust, globally consistent factors, even for narrow, emerging themes. Delivers daily factor returns and per-company price exposures with statistical significance, as regional estimates and global projections.
Best for: thematic return attribution, exposure measurement, signal generation.
Clusters the fundamental factor space into a compact set of 20+ statistical factors at 3-, 6- and 12-month horizons, estimated per region, while preserving interpretability through thematic factor loadings. Correlated themes are resolved into a clean factor structure built for institutional risk workflows.
Best for: risk decomposition, covariance forecasting, regime detection.
Stocks tend to move in response to shared drivers, or “factors”. TFM uniquely breaks down stocks’ returns into thematic and style factors with precision, each model documented in its own methodology. See our documentation below.
Generate thematic factor signals for systematic trading strategies.
Validate investment theses and discover emerging themes. Balance thematic exposures while controlling for style factors.
Monitor and control thematic exposures across global portfolios, with the statistical model's compact factor set built for risk decomposition and covariance forecasting.
Time entry and exit points based on thematic momentum.
