Data Science vs Econometrics: Testing Different Forecasting Approaches
DOI:
https://doi.org/10.55578/jift.2601.003Keywords:
Data Science, Econometrics, Forecast comparisonsAbstract
Forecasting is a critical activity for economists, financial analysts, and businesses engaged in budgeting and planning. Recent advances in Data Science promise greater forecasting sophistication and accessibility, often through automated or semi‑automated tools. However, these techniques also carry risks, particularly when users are unaware of the underlying data‑generating processes. This paper compares popular Data Science forecasting approaches with long‑standing Econometric methods using simulated data under varying assumptions about noise and structural complexity. The results show that while Data Science methods can perform competitively in highly noisy environments, structural econometric models consistently outperform them when data noise is low or when underlying relationships are complex. This reinforces a long‑standing insight often missed by automatic approaches: the source and structure of shocks matter for accurate forecasting.
References
1. Guo, T., & Hauptmann, E. (2024). Fine‑tuning large language models for stock return prediction using newsflow. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (Industry Track), pages 1028–1045.
2. Yan, R., Jin, J., & Han, K. (2024). Reinforcement learning for deep portfolio optimization. Electronic Research Archive, Vol. 32, No. 9, pages 5176–5200.
3. Rahmani, R., Parola, M., & Cimino, M. (2024). A machine learning workflow to address credit default prediction. arXiv. https://arxiv.org/abs/2403.03785
4. Kwan, A., & Philip, R. (2025). Reinforcement learning in a dynamic limit order market. Working paper, University of New South Wales and University of Sydney.
5. Černevičienė, J., & Kabašinskas, A. (2024). Explainable artificial intelligence (XAI) in finance: A systematic literature review. Artificial Intelligence Review, Vol. 57, pages 216–260.
6. Broadbent, B. (2016), “The distributional implications of low structural interest rates and some remarks about monetary policy trade-offs”, Speech given at the Society of Business Economists Annual Conference, 18 November. Available at: https://www.bankofengland.co.uk/-/media/boe/files/speech/2016/the-distributional-implications-of-low-structural-interest-rates.pdf
7. Tüzün, G. (2024), The importance of external shocks and global monetary conditions for a small-open economy: The case of Türkiye, Central Bank Review, Vol. 24, No. 3.
8. Tibshirani, R (1996), “Regression Shrinkage and Selection via the Lasso”, Journal of the Royal Statistical Society, Series B (Methodological) Vol. 58, No. 1, pages 267–88. http://www.jstor.org/stable/2346178.
9. Zou, H. and Hastie, T (2005), “Regularization and Variable Selection Via the Elastic Net”, Journal of the Royal Statistical Society, Series B: Statistical Methodology, Vol. 67, No. 2 (April), pages 301–20. https://doi.org/10.1111/j.1467-9868.2005.00503.x
10. Hendry, D. F. (2024), “A Brief History of General-to-specific Modelling”, Oxford Bulletin of Economics and Statistics, Vol. 86, No. 1, pages 1-20. https://doi.org/10.1111/obes.12578
11. Christiano, L. J. (2012), “Christopher A. Sims and Vector Autoregressions”, Scandinavian Journal of Economics, Vol. 114, No. 4, pages 1082-1104.
12. Uhlig, H. (2005). “What are the effects of monetary policy on output? Results from an agnostic identification procedure”, Journal of Monetary Economics, Vol. 52, No. 2, pages 381–419.
13. Ellis, C., Mumtaz, H. and Zabczyk, P. (2014), “What lies beneath? A time-varying FAVAR model for the UK transmission mechanism”, The Economic Journal, Vol. 124, May, pages 668-699.
14. Ellis, C. and Price, S. (2004), “UK Business Investment and the User Cost of Capital”, The Manchester School, Vol. 72, pages 72-93. https://doi.org/10.1111/j.1467-9957.2004.00420.x
15. Ellis, C. (2006),”Elasticities, markups and technical progress: evidence from a state-space approach”, Bank of England Working Paper No. 300, July.
16. Ellis, C. (2017), “Scenario-Based Stress Tests: Are They Painful Enough?”, Contemporary Economics, Vol. 11, No. 2, pages 219-234.
Downloads
Published
Data Availability Statement
The main data used in this paper are simulated, with the details of the simulations provided within the paper itself.
Issue
Section
License
Copyright (c) 2026 Colin Ellis (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.