Optimal vs naïve diversification in cryptocurrencies

Abstract

This paper contributes to the literature on cryptocurrencies by examining the performance of naïve (1/N) and optimal (Markowitz) diversification in a portfolio of four popular cryptocurrencies. We employ weekly data with weekly rebalancing and show there is very little to select between naïve diversification and optimal diversification. Our results hold for different levels of risk-aversion and an alternative estimation window.

Portfolio management with cryptocurrencies: The role of estimation risk

“…This technique uses alternative estimates for the input parameters and imposes tighter constraints to the weights of assets with higher potential estimation errors and we find that this technique outperforms both 1/N and the Markowitz portfolio optimization framework when applied to a portfolio of cryptocurrencies. Therefore our paper furthers the findings by Platanakis et al (2018a).…”

“…1 The diversification benefits of Bitcoin to other financial assets has been reported by Bouri et al (2017) and Corbet et al (2018a), while recently Kajtazi and Moro (2018) and Platanakis and Urquhart (2018) both report substantial benefits from including Bitcoin in traditional portfolios. Platanakis et al (2018a) show that there is very little to select between optimal mean-variance diversification and 1/N for a portfolio of cryptocurrencies. However, cryptocurrencies have been found to be highly volatile (Chaim and Laurini 2018) and therefore have higher potential estimation errors in their parameters that may make portfolio theory particularly problematic when applied to a portfolio of cryptocurrencies.…”

Portfolio Management with Cryptocurrencies: The Role of Estimation Risk

“…1 The diversification benefits of Bitcoin to other financial assets has been reported by Bouri et al (2017) and Corbet et al (2018a), while recently Kajtazi and Moro (2018) and Platanakis and Urquhart (2018) both report substantial benefits from including Bitcoin in traditional portfolios. Platanakis et al (2018a) show that there is very little to select between optimal mean-variance diversification and 1/N for a portfolio of cryptocurrencies. However, cryptocurrencies have been found to be highly volatile (Chaim and Laurini 2018) and therefore have higher potential estimation errors in their parameters that may make portfolio theory particularly problematic when applied to a portfolio of cryptocurrencies.…”

Technical trading and cryptocurrencies

“…Recently, Platanakis and Urquhart (2019) examine the out-of-sample benefit of including Bitcoin in eight popular asset allocation strategies in portfolios of stocks and bonds. They find that the inclusion of Bitcoin generates substantially higher risk-adjusted returns, where the results are robust to a different structure of estimation windows, the incorporation of transaction costs, the inclusion of a commodity portfolio, an alternative index for Bitcoin as well as two additional portfolio optimization techniques including higher moments with (and without) variance-based constraints.In respect of the optimisation of cryptocurrency portfolios Platanakis et al (2018) examine the performance of nave (1/N) and optimal (Markowitz) diversification in a portfolio of four popular cryptocurrencies. They show there is very little to differentiate between nave diversification and optimal diversification.…”