Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

GENERALIZED INFORMATION CRITERIA FOR SPARSE STATISTICAL JUMP MODELS

Cortese, Federico LU ; Kolm, Petter Nils LU and Lindström, Erik LU orcid (2023) p.68-78
Abstract
We extend the generalized information criteria for high-dimensional penalized
models to sparse statistical jump models, a new class of statistically robust and computationally efficient alternatives to hidden Markov models. In a simulation study, we demonstrate that the new generalized information criteria selects the correct hyperparameters with high probability. Finally, providing an empirical application, we infer the key features that drive the return dynamics of the largest cryptocurrencies. We find that a four-state model best describes the dynamics of cryptocurrency returns. The states have natural market-based interpretations as they correspond to bull, bull-neutral, bear-neutral, and bear market regimes, respectively.
Please use this url to cite or link to this publication:
author
; and
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
host publication
Symposium i anvendt statistik 2023
editor
Linde, Peter
pages
68 - 78
publisher
Danmarks Statistik & Copenhagen Business School
ISBN
978-87-989370-3-6
language
English
LU publication?
yes
id
f86c7854-434c-46c0-8f04-c03c671e5136
alternative location
http://www.statistiksymposium.dk/Symposium%20i%20anvendt%20stalistik%202023_Web.pdf
date added to LUP
2023-04-25 12:28:37
date last changed
2023-05-25 15:14:46
@inbook{f86c7854-434c-46c0-8f04-c03c671e5136,
  abstract     = {{We extend the generalized information criteria for high-dimensional penalized<br/>models to sparse statistical jump models, a new class of statistically robust and computationally efficient alternatives to hidden Markov models. In a simulation study, we demonstrate that the new generalized information criteria selects the correct hyperparameters with high probability. Finally, providing an empirical application, we infer the key features that drive the return dynamics of the largest cryptocurrencies. We find that a four-state model best describes the dynamics of cryptocurrency returns. The states have natural market-based interpretations as they correspond to bull, bull-neutral, bear-neutral, and bear market regimes, respectively.<br/>}},
  author       = {{Cortese, Federico and Kolm, Petter Nils and Lindström, Erik}},
  booktitle    = {{Symposium i anvendt statistik 2023}},
  editor       = {{Linde, Peter}},
  isbn         = {{978-87-989370-3-6}},
  language     = {{eng}},
  pages        = {{68--78}},
  publisher    = {{Danmarks Statistik & Copenhagen Business School}},
  title        = {{GENERALIZED INFORMATION CRITERIA FOR SPARSE STATISTICAL JUMP MODELS}},
  url          = {{http://www.statistiksymposium.dk/Symposium%20i%20anvendt%20stalistik%202023_Web.pdf}},
  year         = {{2023}},
}