Abstract:

Recently cryptocurrency markets have seen an immense growth. Bitcoin is one of the most popular cryptocurrencies accounting for the highest share of all cryptocurrency markets, even though it still remains rather unclear whether it resembles more to a currency, a commodity or an asset. Previous research has shown that Bitcoin is often used for investment purposes, a fact that suggests the importance of analysing its volatility. In this article, we examine the optimal conditional heteroskedasticity model, not only in terms of goodness-of-fit, but also in terms of forecasting performance, an area which has been underexplored in the case of Bitcoin. According to the results, the optimal conditional heteroskedasticity model that can fit the series is not the same as the one that can forecast it better. As modelling GARCH effects in Bitcoin market effectively is crucial for appropriate portfolio management, our results can help investors and other decision makers make more informed decisions.

Keywords: Bitcoin, Cryptocurrencies, GARCH, Volatility, Forecasting

JEL classification: C22, C5, G1

1. Introduction

Over the last few years, the analysis of Bitcoin has drawn a lot of both public and academic attention. Bitcoin is the first implementation of a concept called “cryptocurrency”, which was first described in 1998 by Wei Dai on the cypherpunks mailing list, suggesting the idea of a new form of money that uses cryptography to control its creation and transactions, rather than a central authority, but the first Bitcoin specification was published in 2009 in a cryptography mailing list by Satoshi Nakamoto (Bitcoin.org 2017).

The market of cryptocurrencies has grown remarkably with Bitcoin being considered the most famous cryptocurrency, with an estimated market capitalisation of $19.6 billion (coinmarketcap.com accessed on 8th March 2017), which currently accounts for around 84.4% of the total estimated cryptocurrency capitalisation. An overview of Bitcoin can be found in, e.g., Becker et al. (2013), Dwyer (2015), Frisby (2014), Böhme et al. (2015) and Selgin (2015). Hence, Bitcoin is only briefly introduced here.

It has been previously argued that Bitcoin shares some elements of currencies. However, recent fluctuations in Bitcoin prices (see Figure 1) have resulted in unpredictable volatility undermining the role Bitcoin plays as a unit of account (Cheah and Fry 2015), while users have adopted Bitcoin not only as a currency but also for investment purposes. In fact, new users tend to trade Bitcoin on a speculative investment intention basis and have low intention to rely on the underlying network as means for paying goods or services (Glaser et al. 2014). The Bitcoin market is thus highly speculative at present, and therefore Bitcoin may be mostly used as an asset rather than a currency (Baek and Elbeck 2015; Dyhrberg 2016a).