Diverging Roads

Theory-Based vs. Machine Learning-Implied Stock Risk Premia

authored by: Joachim Grammig, Constantin Hanenberg, Christian Schlag, Jantje Sönksen
Abstract: We assess financial theory-based and machine learning methods to quantify stock risk premia and investigate the potential of hybrid strategies. In the low signal-to-noise environment of a one-month investment horizon, we recommend to rely on a theory-based strategy that exploits the information in current option prices, especially if the risk premium estimate is to be updated at a high frequency. At the one-year horizon, the theory/option-based strategy and an ensemble of neural networks, two notably different methodologies, perform comparably well. A random forest can improve on the theory-based method, provided that a sufficiently long training period is used. In an effort to connect the opposing philosophies, we identify the use of a random forest to account for the approximation errors of the theory-based approach towards measuring stock risk premia as a promising hybrid strategy. It combines the advantages of two diverging roads in the finance world.
Organisation(s): Institute of Econometrics and Data Science
External Organisation(s): University of Tübingen
University of Cologne
Goethe University Frankfurt
Leibniz Institute for Financial Research SAFE
Type: Article
Journal: Journal of Financial Econometrics
Volume: 23
Pages: 1-55
No. of pages: 90
ISSN: 1479-8409
Publication date: 10.03.2025
Publication status: Published
Peer reviewed: Yes
Electronic version(s): https://doi.org/10.1093/jjfinec/nbaf005 (Access: Closed)
https://doi.org/10.2139/ssrn.3536835 (Access: Open)
https://doi.org/10.15496/publikation-39286 (Access: Open)
: Details in the research portal "Research@Leibniz University"

BibTeX

@article{22969132ae554db5b1d89108bd3ff6e2,
title = "Diverging Roads: Theory-Based vs. Machine Learning-Implied Stock Risk Premia",
abstract = "We assess financial theory-based and machine learning methods to quantify stock risk premia and investigate the potential of hybrid strategies. In the low signal-to-noise environment of a one-month investment horizon, we recommend to rely on a theory-based strategy that exploits the information in current option prices, especially if the risk premium estimate is to be updated at a high frequency. At the one-year horizon, the theory/option-based strategy and an ensemble of neural networks, two notably different methodologies, perform comparably well. A random forest can improve on the theory-based method, provided that a sufficiently long training period is used. In an effort to connect the opposing philosophies, we identify the use of a random forest to account for the approximation errors of the theory-based approach towards measuring stock risk premia as a promising hybrid strategy. It combines the advantages of two diverging roads in the finance world.",
keywords = "diverging, machine learning, risk, premia, stock risk",
author = "Joachim Grammig and Constantin Hanenberg and Christian Schlag and Jantje S{\"o}nksen",
year = "2025",
month = mar,
day = "10",
doi = "10.1093/jjfinec/nbaf005",
language = "English",
volume = "23",
pages = "1--55",
journal = "Journal of Financial Econometrics",
issn = "1479-8409",
publisher = "Oxford University Press",
number = "2",
}