Dynamic Nelson Siegel Model with Time-Varying Neural Factor Loadings
Kooiker, S., van Brummelen, J., Schaumburg, J., & Zamojski, M. (2026). Dynamic Nelson Siegel Model with Time-Varying Neural Factor Loadings. Working Paper.
CV
Education
- Ph.D. in Econometrics, Vrije Universiteit Amsterdam, The Netherlands, 2023 - present
- Focus on developing machine learning models for forecasting in time-varying conditions
- Presentations at ISF Dijon 2024, CFE London 2024, IAAE Torino 2025, ISF Beijing 2025, FinEML Rotterdam 2025, ECB Frankfurt 2026
- M.Sc. Artificial Intelligence, Cum Laude, Vrije Universiteit Amsterdam, The Netherlands, 2022 - 2024
- M.Sc. Econometrics and Data Science, Vrije Universiteit Amsterdam, The Netherlands, 2021 - 2023
- B.Sc. Econometrics and Operations Research, Vrije Universiteit Amsterdam, The Netherlands, 2018 - 2021
- Summer School on Financial Machine Learning, Yale University, New Haven, USA, 2025
- Summer School in C++, Jülich Supercomputing Centre, Germany, 2025
- Courses at Tinbergen Institute, MPhil graduate school, The Netherlands, 2023
Employment
- September 2023 - present: Ph.D. Candidate in Econometrics
- Vrije Universiteit Amsterdam, School of Business and Economics
- Developing machine learning models for forecasting in time-varying conditions
- Coordinating two 600+ student courses
- Supervising students during their thesis projects
- Presented papers at multiple international conferences
- 2023: External Expert
- European Commission, European Union
- Investigated how Large Language Models affect secondary education
- 2022 - 2023: Research Assistant
- Vrije Universiteit Amsterdam, School of Business and Economics
- Working on Growth-at-Risk forecasting with Neural Networks
- Presentation at The Netherlands Econometric Study Group
- Collaborating with senior researchers
- 2021 - 2022: Risk Analyst
- RiskQuest, Amsterdam, The Netherlands
- Contributed to financial risk assessment for several banks and hedge funds
- Performed research in credit rating transition forecasting
Research Fields
Time Series Econometrics, Forecasting, Machine Learning, Time-Varying Methods, and Macroeconometrics
IT Skills
- Data Analytics
- Extensive experience developing and implementing algorithms for data analytics using Python, R, Matlab and Julia
- Some knowledge of C++
- In-depth knowledge about using programming software for statistical learning
- Expert on working with High Performance Computing for Econometrics
- Support Tools
- Experience with Office 365 including programming of VBA macros
- Experience with working with LaTeX
- Proficient in using Git for version control and collaborative development
- Integration of Large Language Models, online and via API
Publications
Nonparametric Time-Varying Granger Causality
Kooiker, S. (2025). Nonparametric Time-Varying Granger Causality. Working Paper.
Multi-period Growth-at-Risk Forecasting with Recurrent Neural Network
Kooiker, S., Hoesch, L., & Schaumburg, J. (2024). Multi-period Growth-at-Risk Forecasting with Recurrent Neural Network. Working Paper.
Talks
Dynamic Nelson Siegel Model with Time-Varying Neural Factor Loadings
Conference Presentation at ECB Conference on Forecasting Techniques, Frankfurt, Germany
Dynamic Nelson Siegel Model with Time-Varying Neural Factor Loadings
Conference Presentation at ISF 2025 Conference in Forecasting, Beijing, China
Nonparametric Time-Varying Granger Causality
Conference Presentation at IAAE 2025 Conference in Applied Econometrics, Turin, Italy
Dynamic Nelson Siegel Model with Time-Varying Neural Factor Loadings
Conference Presentation at FinEML Conference in Financial Econometrics and Machine Learning, Rotterdam, Netherlands
Econometric Game 2025 Finalist
Competition at Econometric Game 2025, Amsterdam, Netherlands
Nonparametric Time-Varying Granger Causality
Conference Presentation at CFE-CMS 2024 Conference in Financial Econometrics and Statistics, London, United Kingdom
Multi-period Growth-at-Risk Forecasting with Recurrent Neural Network
Conference Presentation at ISF 2024 Conference in Forecasting, Dijon, France
Multi-period Growth-at-Risk Forecasting with Recurrent Neural Network
Poster Presentation at National Econometrics Study Group, Rotterdam, Netherlands
Multi-period Growth-at-Risk Forecasting with Recurrent Neural Network
Workshop Presentation at 11th SIdE-IEA Workshop for PhD Students in Econometrics and Empirical Economics, Bertinoro, Italy
Teaching
Service and Leadership
- Econometric Game 2025 Finalist, University of Amsterdam, Amsterdam, Netherlands, 2025
- Three-day paper on grid congestion forecasting in Germany
- National Econometrics Day Committee, Study Association Kraket, Amsterdam, The Netherlands, 2020 - 2021
- Together with a team of five fellow Econometricians, organized the National Econometrics Day for 800+ Econometrics students from all over the Netherlands