Premium
A hidden Markov model to detect regime changes in cryptoasset markets
Author(s) -
Giudici Paolo,
Abu Hashish Iman
Publication year - 2020
Publication title -
quality and reliability engineering international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.913
H-Index - 62
eISSN - 1099-1638
pISSN - 0748-8017
DOI - 10.1002/qre.2673
Subject(s) - econometrics , markov chain , covariance , covariance matrix , cryptocurrency , hidden markov model , economics , markov model , computer science , mathematics , statistics , artificial intelligence , algorithm , machine learning , computer security
The objective of this work is to understand the dynamics of cryptocurrency prices. Specifically, how prices switch between different regimes, going from “bull” to “stable” and “bear” times. For this purpose, we propose a hidden Markov model that aims at explaining the evolution of Bitcoin prices through different, unobserved states. The implementation of the proposed model includes a likelihood ratio test that allows to compare models with different states and with different covariance structures. Our empirical findings show that the time movements of Bitcoin prices across different exchange markets are well‐described by the proposed model. In particular, a parsimonious model with a diagonal covariance matrix leads to better predictions, compared with a model with a full covariance matrix.