Please use this identifier to cite or link to this item:
|Title:||True or spurious long memory in the cryptocurrency markets: evidence from a multivariate test and other Whittle estimation methods||Authors:||Assaf, Ata
Gil-Alana, Luis Alberiko
|Affiliations:||Department of Business Administration||Keywords:||Cryptocurrency markets
Multivariate long-memory tests
Spurious long memory
|Issue Date:||2022||Part of:||Empirical economics||Volume:||63||Issue:||3||Start page:||1543||End page:||1570||Abstract:||
This paper applies a new proposed multivariate score-type test against spurious long memory to a group of cryptocurrency market returns. The test statistic developed by Sibbertsen et al. (J Econ 203(1): 33–49, 2018) is based on the multivariate local Whittle likelihood function and is proven to be consistent against the alternative two cases of random level shifts and smooth trends. We apply the test to the returns, absolute returns, and modified absolute returns. Overall, the recently developed test statistic fails to reject the null hypothesis of true long memory for most cryptocurrencies, except for the Stellar market. Therefore, applying the new test statistic supports the argument that the long memory in the cryptocurrency markets is real and is not a spurious one. Our results are further supported by applying other consistent local Whittle methods that allow for the estimation of the memory parameter by accounting for the presence of perturbations or low-frequency contaminations.
|URI:||https://scholarhub.balamand.edu.lb/handle/uob/5374||ISSN:||03777332||DOI:||10.1007/s00181-021-02165-6||Ezproxy URL:||Link to full text||Type:||Journal Article|
|Appears in Collections:||Department of Business Administration|
Show full item record
checked on Aug 20, 2022
checked on Feb 5, 2023
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.