Folded Bivariate Distributions as Models for Magnitude Correlation

Emmanuel Afuecheta; Saralees Nadarajah; Stephen Chan

doi:10.57805/revstat.v21i1.395

Folded Bivariate Distributions as Models for Magnitude Correlation

Authors

Emmanuel Afuecheta King Fahd University of Petroleum & Minerals https://orcid.org/0000-0002-9223-0799
Saralees Nadarajah University of Manchester https://orcid.org/0000-0002-0481-0372
Stephen Chan American University of Sharjah https://orcid.org/0000-0002-2312-2137

DOI:

https://doi.org/10.57805/revstat.v21i1.395

Keywords:

estimation, magnitude correlation of stock returns, value at risk

Abstract

The concept of magnitude correlation requires the use of folded bivariate distributions. However, apart from the folded bivariate normal and folded bivariate t distributions (of these two only the former has received any real applications), nothing is known about folded bivariate distributions. Here, we introduce six new folded bivariate distributions. Applications involving stock indices of ten major economies show the value of the proposed distributions.

Downloads

REVSTAT2023v21n1_2

Published

2023-05-26

How to Cite

Afuecheta , E., Nadarajah , S., & Chan , S. (2023). Folded Bivariate Distributions as Models for Magnitude Correlation. REVSTAT-Statistical Journal, 21(1), 21–38. https://doi.org/10.57805/revstat.v21i1.395