Beyond Bitcoin: A Statistical Comparison of Leading Cryptocurrencies and Fiat Currencies and Their Impact on Portfolio Diversification

Abstract

Cryptocurrencies have developed very dynamically although their future role is yet unclear. In any event, they are too big to ignore. The purpose of this article is to contribute to the understanding of cryptocurrencies in an individual and in a portfolio context. The study is based on daily closing prices of leading cryptocurrencies (Bitcoin, Ethereum, Ripple, Litecoin, and Dash) and fiat currencies (EUR, GBP, CHF, CAD, and JPY), all measured against USD. The analysis is threefold: First, the authors analyze basic statistical properties, such as correlation and autocorrelation of returns. Second, they perform a Kolmogorov–Smirnov test (KS test) and a variance ratio test (VRT) with heteroscedasticity adjustment. Third, they solve more than 4,800 optimization problems to analyze the impact of individual crypto- and fiat currencies on portfolio diversification. Among other findings, the authors find that Bitcoin, Ethereum, Dash, CAD, JPY, and EUR contribute most to reduce the variance of a mixed portfolio. In a portfolio consisting of cryptocurrencies only, Bitcoin and Ripple have the largest diversification effect. The findings provide insights for investors who focus on minimum variance portfolios or, more generally, for investors who seek to reduce return volatility exposure, as well as for monetary authorities, cryptocurrency issuers, and providers of market infrastructure. TOPICS: Real assets/alternative investments/private equity, statistical methods, portfolio construction