Dynamic efficiency and arbitrage potential in Bitcoin: A long-memory approach

Kun Duan; Zeming Li; Andrew Urquhart; Jinqiang Ye

doi:10.1016/j.irfa.2021.101725

Dynamic efficiency and arbitrage potential in Bitcoin: A long-memory approach

Kun Duan, Zeming Li, Andrew Urquhart^*, Jinqiang Ye

^*Corresponding author for this work

Finance

Research output: Contribution to journal › Article › peer-review

Abstract

Employing a long-memory approach, we provide a study of the evolution of informational efficiency in five major Bitcoin markets and its influence on cross-market arbitrage. While all the markets are close to full informational efficiency over the whole sample period, the degree of market efficiency varies across markets and over time. The cross-market discrepancy in market efficiency gradually vanishes, suggesting the segmented markets are developing to a consensus where all markets are equally efficient. Through a fractionally cointegrated vector autoregressive (FCVAR) model we show that when the efficiency in Bitcoin/USD and Bitcoin/AUD markets improves the cross-market arbitrage potential narrows, whereas it widens when the efficiency in Bitcoin/CAD, Bitcoin/EUR, and Bitcoin/GBP markets improves. A battery of robustness checks reassure our main findings.

Original language	English
Article number	101725
Number of pages	19
Journal	International Review of Financial Analysis
Volume	75
Early online date	26 Mar 2021
DOIs	https://doi.org/10.1016/j.irfa.2021.101725
Publication status	Published - May 2021

Keywords

Bitcoin
Market efficiency
Cryptocurrency
Long memory
FCVAR

Access to Document

10.1016/j.irfa.2021.101725Licence: None: All rights reserved

Cite this

@article{83f76191b1d04995bcf17934c52dad80,

title = "Dynamic efficiency and arbitrage potential in Bitcoin: A long-memory approach",

abstract = "Employing a long-memory approach, we provide a study of the evolution of informational efficiency in five major Bitcoin markets and its influence on cross-market arbitrage. While all the markets are close to full informational efficiency over the whole sample period, the degree of market efficiency varies across markets and over time. The cross-market discrepancy in market efficiency gradually vanishes, suggesting the segmented markets are developing to a consensus where all markets are equally efficient. Through a fractionally cointegrated vector autoregressive (FCVAR) model we show that when the efficiency in Bitcoin/USD and Bitcoin/AUD markets improves the cross-market arbitrage potential narrows, whereas it widens when the efficiency in Bitcoin/CAD, Bitcoin/EUR, and Bitcoin/GBP markets improves. A battery of robustness checks reassure our main findings.",

keywords = "Bitcoin, Market efficiency, Cryptocurrency, Long memory, FCVAR",

author = "Kun Duan and Zeming Li and Andrew Urquhart and Jinqiang Ye",

year = "2021",

month = may,

doi = "10.1016/j.irfa.2021.101725",

language = "English",

volume = "75",

journal = "International Review of Financial Analysis",

issn = "1057-5219",

publisher = "Elsevier",

}

TY - JOUR

T1 - Dynamic efficiency and arbitrage potential in Bitcoin

T2 - A long-memory approach

AU - Duan, Kun

AU - Li, Zeming

AU - Urquhart, Andrew

AU - Ye, Jinqiang

PY - 2021/5

Y1 - 2021/5

N2 - Employing a long-memory approach, we provide a study of the evolution of informational efficiency in five major Bitcoin markets and its influence on cross-market arbitrage. While all the markets are close to full informational efficiency over the whole sample period, the degree of market efficiency varies across markets and over time. The cross-market discrepancy in market efficiency gradually vanishes, suggesting the segmented markets are developing to a consensus where all markets are equally efficient. Through a fractionally cointegrated vector autoregressive (FCVAR) model we show that when the efficiency in Bitcoin/USD and Bitcoin/AUD markets improves the cross-market arbitrage potential narrows, whereas it widens when the efficiency in Bitcoin/CAD, Bitcoin/EUR, and Bitcoin/GBP markets improves. A battery of robustness checks reassure our main findings.

AB - Employing a long-memory approach, we provide a study of the evolution of informational efficiency in five major Bitcoin markets and its influence on cross-market arbitrage. While all the markets are close to full informational efficiency over the whole sample period, the degree of market efficiency varies across markets and over time. The cross-market discrepancy in market efficiency gradually vanishes, suggesting the segmented markets are developing to a consensus where all markets are equally efficient. Through a fractionally cointegrated vector autoregressive (FCVAR) model we show that when the efficiency in Bitcoin/USD and Bitcoin/AUD markets improves the cross-market arbitrage potential narrows, whereas it widens when the efficiency in Bitcoin/CAD, Bitcoin/EUR, and Bitcoin/GBP markets improves. A battery of robustness checks reassure our main findings.

KW - Bitcoin

KW - Market efficiency

KW - Cryptocurrency

KW - Long memory

KW - FCVAR

U2 - 10.1016/j.irfa.2021.101725

DO - 10.1016/j.irfa.2021.101725

M3 - Article

SN - 1057-5219

VL - 75

JO - International Review of Financial Analysis

JF - International Review of Financial Analysis

M1 - 101725

ER -

Dynamic efficiency and arbitrage potential in Bitcoin: A long-memory approach

Abstract

Keywords

Access to Document

Fingerprint

Cite this