Self-driving neural networks for term structure modeling

Authors: Sicco Kooiker, Janneke van Brummelen, Julia Schaumburg, and Marcin Zamojski

Abstract:

We propose a factor model with time-varying loadings for term structure modeling and fore- casting. While maintaining the interpretation of the factors as level, slope, and curvature through explicit identification restrictions, we allow the loadings to take flexible shapes by specifying them as neural networks that evolve over time using a “self-driving” updating scheme based on past forecast errors, with gradient scaling to improve robustness. Using an empirically calibrated simulation study and an application to U.S. Treasury yields across 24 maturities, we show that flexible and dynamic factor loadings improve forecasting per- formance relative to standard benchmarks, including Nelson–Siegel models and the random walk. The gains are strongest at medium maturities and shorter forecast horizons, highlight- ing the importance of capturing curvature dynamics. In-sample results further illustrate how time-varying loadings provide insight into changes in yield curve shape beyond traditional parametric specifications.

Presentations:

• ECB Conference on Forecasting Techniques, Frankfurt, Germany (2026)
• FinEML Conference in Financial Econometrics and Machine Learning, Rotterdam, Netherlands (2025)
• ISF Conference, Beijing, China (2025)

Recommended citation: Kooiker, S., van Brummelen, J., Schaumburg, J., & Zamojski, M. (2026). Self-Driving Neural Networks for Term Structure Modeling. Working Paper.
Download Paper