Multifractal Volatility Models: Beyond the Binary Case
Journal of Financial Econometrics · 17.09.2026 · 24 (5)
DOI: 10.1093/jjfinec/nbag030 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Financial Econometrics, online erschienen am 05.11.2026. Verfasst von Maciej Augustyniak und 2 weiteren Autor:innen. Die Arbeit nennt 67 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Natural Sciences and Engineering Research Council of Canada, Fonds de recherche du Québec—Nature et technologies.
Abstract (Originalsprache)
Factorial hidden Markov processes, which build volatility from independent two-state Markov chains, have become strong competitors to GARCH models. We show that they can be represented as multi-component stochastic volatility models with non-geometric sojourn times, providing a theoretical basis for their empirical performance. We introduce the Multifractal Discrete Stochastic Volatility (MDSV) process, which allows more than two states per component and a more flexible stationary distribution, and develop a filtering algorithm that substantially reduces the cost of likelihood evaluation. On VOLARE realized variances, the MDSV process delivers more accurate density and volatility tail-risk forecasts than log-HAR and ARFIMA benchmarks.
Förderung
- Natural Sciences and Engineering Research Council of Canada
- Fonds de recherche du Québec—Nature et technologies
Zitieren
APA
Augustyniak, M., Dufays, A., & Maoude, K. A. H. (2026). Multifractal Volatility Models: Beyond the Binary Case. Journal of Financial Econometrics, 24(5). https://doi.org/10.1093/jjfinec/nbag030
Vancouver
Augustyniak M, Dufays A, Maoude KAH. Multifractal Volatility Models: Beyond the Binary Case. Journal of Financial Econometrics. 2026;24(5). doi:10.1093/jjfinec/nbag030
BibTeX
@article{Augustyniak2026,
author = {Augustyniak, Maciej and Dufays, Arnaud and Maoude, Kassimou Abdoul Haki},
title = {Multifractal Volatility Models: Beyond the Binary Case},
journal = {Journal of Financial Econometrics},
year = {2026},
volume = {24},
number = {5},
doi = {10.1093/jjfinec/nbag030}
}
EUVIMED