BITCOIN RETURN PREDICTABILITY, GLOBAL FINANCIAL DETERMINANTS, AND RISK MANAGEMENT: EVIDENCE FROM MACHINE LEARNING
Advances and Applications in Statistics · 2026 · 93 (11) · 1355-1391
DOI: 10.17654/0972361726061 ↗Kurz gefasst
Fachartikel in der Zeitschrift Advances and Applications in Statistics, 2026. Verfasst von Nadia Belkhir.
Abstract (Originalsprache)
This study investigates the relationship between Bitcoin and major global financial indicators, including Ethereum, gold prices, WTI crude oil, the S&P 500 AI index, the Nasdaq index, and the VIX index. Using daily data from January 2019 to July 2025, several machine-learning models generated through the \(\mathrm{H}_{2} \mathrm{O}\) AutoML framework were applied to forecast Bitcoin returns. The results indicate that the deep learning model achieved the lowest out-of-sample prediction error, followed by the GLM and stacked ensemble > models, highlighting their relative predictive performance in forecasting Bitcoin returns. In addition, Individual Conditional Expectation (ICE) analysis reveals heterogeneous effects of global financial variables on Bitcoin price movements. Overall, the findings demonstrate that deep neural networks and automated machine-learning frameworks provide powerful tools for predicting cryptocurrency prices in highly volatile financial markets.
Zitieren
APA
Belkhir, N. (2026). BITCOIN RETURN PREDICTABILITY, GLOBAL FINANCIAL DETERMINANTS, AND RISK MANAGEMENT: EVIDENCE FROM MACHINE LEARNING. Advances and Applications in Statistics, 93(11), 1355-1391. https://doi.org/10.17654/0972361726061
Vancouver
Belkhir N. BITCOIN RETURN PREDICTABILITY, GLOBAL FINANCIAL DETERMINANTS, AND RISK MANAGEMENT: EVIDENCE FROM MACHINE LEARNING. Advances and Applications in Statistics. 2026;93(11):1355-1391. doi:10.17654/0972361726061
BibTeX
@article{Belkhir2026,
author = {Belkhir, Nadia},
title = {BITCOIN RETURN PREDICTABILITY, GLOBAL FINANCIAL DETERMINANTS, AND RISK MANAGEMENT: EVIDENCE FROM MACHINE LEARNING},
journal = {Advances and Applications in Statistics},
year = {2026},
volume = {93},
number = {11},
pages = {1355--1391},
doi = {10.17654/0972361726061}
}
EUVIMED