Correlated at the Tail: Implications of Asymmetric Tail-Dependence Across Bitcoin Markets
- PDF / 2,420,361 Bytes
- 11 Pages / 439.37 x 666.142 pts Page_size
- 42 Downloads / 197 Views
Correlated at the Tail: Implications of Asymmetric Tail‑Dependence Across Bitcoin Markets Stelios Bekiros1 · Axel Hedström2 · Evgeniia Jayasekera3 · Tapas Mishra4 · Gazi Salah Uddin2 Accepted: 4 October 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This paper is the first to fully characterize the relationship among cross-market Bitcoin prices to provide a complete picture of directional predictability of Bitcoin traded in various currencies across five developed markets. To exploit full-distributional dynamics, we employ Cross-quantilogram based Correlation and Dependence model to delve deep into the estimates an asymmetric tail dependence across quantiles would reflect on heterogeneous movement pattern of Bitcoin prices. A crossquantilogram-based analysis reveals new empirical evidence of a heterogeneous tail dependence pattern: whereas Bitcoin-USD and the Northeast Asian market (viz., Japan) depicts a strong co-movement, smaller markets display weak connectedness and strong market-efficiency. Keywords Cross-quantilogram · Cross-market Bitcoin prices · Time-varying stability
* Stelios Bekiros [email protected] Axel Hedström [email protected] Evgeniia Jayasekera [email protected] Tapas Mishra [email protected] Gazi Salah Uddin [email protected] 1
European University Institute, Florence, Italy
2
Department of Management and Engineering, Linköping University, Linköping, Sweden
3
National College of Ireland, Dublin, Ireland
4
Southampton Business School, University of Southampton, Southampton, UK
13
Vol.:(0123456789)
S. Bekiros et al.
1 Introduction Predicting Cryptocurrency price movements is one of the most challenging tasks an investor would embark on—thanks to the lack of a strong asset pricing theory, dominance of a largely unregulated market structure, and instrumental role of an implicit value of investors’ sentiments. Recent research in this regard demonstrates that a large part of variance in a cryptocurrency market, such as Bitcoin, is due to the sensitivity of investors to macroeconomic performances of a country (e.g., Corbet et al. 2018; Cheah et al. 2018; Gillaizeau et al. 2019). A common practice is to treat observed pattern of price movement as an ‘information set’, which an investor uses to ‘predict’ as his next strategy of investment. But, this information set conceals unaccounted for noisy signals arising out of, for instance, dynamic movements in macroeconomic fundamentals (representing economic parameter-driven sentimental values) from other markets. Eventually, a component of this ‘information set’ specific to a market, becomes a common component across other markets, because noises generally display transmissive and transformative effects (Gillaizeau et al. 2019). The problem most often neglected is that whilst it is the entire dynamic path of a cryptocurrency price and associated factors that determine the information set, but inference is based only on the centre of the distribution. There are essentially two ways t
Data Loading...
