Constructing cointegrated cryptocurrency portfolios for statistical arbitrage

Abstract

Purpose

This paper aims to present a methodology for constructing cointegrated portfolios consisting of different cryptocurrencies and examines the performance of a number of trading strategies for the cryptocurrency portfolios.

Design/methodology/approach

The authors apply a series of statistical methods, including the Johansen test and Engle–Granger test, to derive a linear combination of cryptocurrencies that form a mean-reverting portfolio. Trading systems are designed and different trading strategies with stop-loss constraints are tested and compared according to a set of performance metrics.

Findings

The paper finds cointegrated portfolios involving four cryptocurrencies: Bitcoin (BTC), Ethereum (ETH), Bitcoin Cash (BCH) and Litecoin (LTC), and the corresponding trading strategies are shown to be profitable under different configurations.

Originality/value

The main contributions of the study are the use of multiple altcoins in addition to bitcoin to construct a cointegrated portfolio, and the detailed comparison of the performance of different trading strategies with and without stop-loss constraints.

Citation

Leung, T. and Nguyen, H. (2019), "Constructing cointegrated cryptocurrency portfolios for statistical arbitrage", Studies in Economics and Finance, Vol. 36 No. 4, pp. 581-599. https://doi.org/10.1108/SEF-08-2018-0264