Essays on memory and dynamics of connectedness in bitcoin markets

Abstract

This thesis sheds new light on the exogenous and endogenous determinants of volatility in Bitcoin prices across many major countries around the globe. Different empirical strategies are proposed to investigate and understand the complex behaviour of volatility, its movements and significant persistence. Chapter Two identifies and characterises the ‘givers and receivers' of volatility in cross-market Bitcoin prices and discusses international diversification strategies in this context. Using both time and frequency domain mechanisms, we provide estimates of outward and inward spillover effects. These have implications for (weak-form) cross-market inefficiency. In our setting, we treat a high degree of spillover as an indicator of weak-form inefficiency, because investors can utilise information on the dynamic spillover effects to produce best long-run predictions of the market. Our results show that Bitcoin prices depict strong (dynamic) spillover in volatility, especially during episodes of high uncertainty. The Bitcoin-USD exchange rate possesses net predictive power, mirrored by the tendency of the Bitcoin-EURO market as a net receiver relative to other markets. Robustness exercise generally supports our claim. The overall implication is that during episodes of high uncertainty, Bitcoin markets depict greater dynamic inefficiency, instrumenting the role of asymmetric information in the path-dependence and predictive power of Bitcoin prices in an interdependent market.

Chapter Three investigates the endogenous growth mechanisms of Bitcoin prices aligned with empirical tests designed to show whether persistence is a product of such a model. However, characterising learning in the Bitcoin market is exceedingly complex, as it is frequently affected by news and/or economic/financial dynamics. Sudden arrival of a shock (for instance, Brexit) can break the cycle of endogenous persistence generating mechanisms. We propose a variant of ARFIMA Markov Switching, with endogenous switch governing the internal dynamics of Bitcoin prices or volatility system. This MS-ARFIMA (endogenous) is synchronised with different mechanisms and shows the credible role of policy on containing volatility persistence. Our model and empirical strategies are new, and our results show the significance of true memory under episodes of structural breaks.

Chapter Four studies Bitcoin prices/volatility during cyber attacks and identifies how they can be seriously manipulated in some markets. In the meantime, exchange rate differentials across markets offer investors the opportunity to enhance portfolio returns. Under these scenarios, it is expected that price volatility on one particular Bitcoin-to-currency exchange market (e.g. Bitcoin-USD) can flow to other markets and can also be acquired from others. Any quantitative information on the centrality or relative isolation of some Bitcoin-to-currency markets can actually help investors to better anticipate their complex dynamic behaviour and exploit potential for forecast-able gain. These premises are rigorously tested in the current paper, using daily price data on six major Bitcoin-to-currency exchange rates. We show the net predictive power and the net receiver of volatility during different cyber attacks. Eventually, such tendencies could help investors design trending strategies to systematically beat the market, hedging and diversifying their investment to maximise profit with the lowest associated risk, and speculating on the behaviour of the market in future attacks.

More information

Identifiers

ORCID for Ahmad Maaitah: orcid.org/0000-0001-6515-5335

ORCID for Tapas Mishra: orcid.org/0000-0002-6902-2326

Catalogue record

Export record