Global portfolio diversification with cryptocurrencies
par Salma Ouali
Université de Neuchâtel - Master of science in finance 2019
Despite the global financial instability, the strong past financial performance of bitcoin (BTC) made it comparable to a digital gold. Therefore, a strand of studies examined Bitcoin and its properties as an asset class. Yermack et al. (2013) compare BTC's daily exchange rates with gold and fiat currencies and argue that BTC is more comparable to a speculative investment. Baur et al. (2015) investigate statistical properties and future usage of bitcoin. They suggest that bitcoin is a hybrid between conventional currencies and commodity currency. Dyhrberg et al. (2015) find similarities of BTC with gold and USD and confirm the views of Baur et al. (2015). Ciaian, Rajcaniova and Kancs (2016) go a step further and analyze the impact of BTC's supply and demand on its price formation. Thus, they discover similar price formation patterns with other currencies.
Whether considered as a medium of exchange or a highly speculative asset, BTC's skyrocketing returns are the main reason of its attractiveness. After being accepted as a method of payment, its use as a financial investment vehicle has been thoroughly examined. In point of fact, a broad
range of literature studied whether including bitcoin in a traditional portfolio can enhance its return and reduce the risk.
Brière et al. (2015), using mean variance spanning test, assert BTC's diversification benefits in an investment portfolio. They conclude that BTC's low correlation with other assets enhances the portfolio's performance and compensates for the increase in the overall risk. As returns of BTC exhibit high kurtosis and low skewness, Eisl et al. (2015), proposes the use of a CVaR framework as an alternative to mean variance analysis. Their findings are in line with the ones of Brière et al. (2015). Indeed, investing a small fraction in BTC enhances intensely the portfolio's risk return tradeoff.
Kihoon Hang (2016), performs a trading strategy based on momentum (TSMOM). By conducting a mean variance analysis, he demonstrates that enhanced returns and reduced volatility can be achieved in a portfolio of equities and TSMOM.
Diversification and hedging are characteristics of huge importance when building a portfolio. Therefore, an investigation of correlation between BTC and other financial assets is crucial. Following this, Bouri et al. (2017), employed a dynamic conditional correlation model (Engle, 2002) to estimate correlation between BTC and other financial assets. They find that Bitcoin does qualify as an effective diversifier. However, it can only serve as a hedge or safe haven in few cases. On the contrary, Bouoiyour and Selmi (2017), vindicate the hedge and safe haven role of bitcoin against oil price movements. Moreover, by adding Bitcoin to a portfolio of oil only, they observe an enhancement in performance and downside risk reduction.
Guesmi et al. (2018), consider DCC GJR GARCH as the suitable model for modeling joint dynamics of different financial assets. They document that Bitcoin can be a systematic hedge and that adding a small fraction of bitcoin reduces considerably the portfolio's risk exposure.
They also stress that a short position in Bitcoin hedges the risk for commodities, stocks and currencies.
Due to the high trading and predominant mining activity in China, Kajtazi and Moro (2017), raise the question of whether BTC adds value in a Chinese asset-based portfolio. They show significant but low correlation with Chinese assets and argue that Bitcoin fails in enhancing portfolio's performance over all periods.
Evidence that Bitcoin offers diversification benefits in a traditional portfolio is abundant in literature and so far, research towards altcoins remains infrequent.
Aiming to close this gap, Osterrieder et al. (2016), were amongst the first to explore the interdependences between cryptocurrencies (CCs). They demonstrate that low mildly correlation exist between most CCs except for the ones sharing the same technology. Following this, Braineis and Mestel (2018), expand the research by exploring the effects of diversified cryptocurrency investments. Their empirical results show that adding several CCs expands the efficient frontier of a bitcoin-based portfolio.
Further investigating the diversification effect, Chuen et al. (2017), advocate the use of the CRIX index, which was developed by Trimborn and Hardle (2016) to mimic the CCs market performance. When looking at static and dynamic correlations between CRIX and traditional assets, they give evidence of diversification benefits. However, the mean-variance spanning tests reveal enhancement of global minimum variance only. In fact, the negative skewness of CRIX and high risk makes it redundant in a tangency portfolio.
Klein et al. (2018), argue that Bitcoin is not the new gold. According to their research results, Bitcoin is only suitable for diversification benefits and does not display stable hedging
capabilities. Meanwhile, they find CRIX to display better hedging effects. Yet, it still fails to be an effective hedge like gold.