David Adugh Kuhe*, Enobong Francis Udoumoh and Moses Tarfa Ujah
Department of Statistics, Joseph Sarwuan Tarka University, Makurdi-Nigeria
*Corresponding author’s Email: davidkuhe@gmail.com, doi.org/10.55639/607.111009
ABSTRACT
This study aims to evaluate optimal cryptocurrencies for both short term and long-term investment through the use of volatility mean reversion and half-life models. The study analyzes the volatility of daily closing prices for twelve cryptocurrencies: Bitcoin, Ethereum, Cardano, Dogecoin, Litecoin, Binance, Ripple, Bitcoin Cash, Stellar, Chainlink, Monero, and Tether over the period from January 14, 2014, to July 16, 2021. The study employs various statistical methods, including summary statistics, normality measures, Augmented Dickey-Fuller unit root test, heteroskedasticity test for ARCH effects, and symmetric GARCH (1,1) models. Findings reveal non-Gaussian, leptokurtic log returns with volatility clustering and high shock persistence across cryptocurrency returns, indicating stable, mean reverting, and predictable log returns. The study found that all cryptocurrencies exhibit mean reversion to their historical mean values after specific periods. Dogecoin, Litecoin, Ripple, Monero, Ethereum, Cardano, Bitcoin Cash, Chainlink, and Binance exhibit faster mean reversion and smaller volatility half-lives of 2, 5, 8, 10, 13, 14, 22, 24, and 35 days respectively, while Bitcoin and Stellar have slower mean reversion rates and larger volatility half-lives of 1454 and 624 days respectively. The study suggests short-term trading and investment in Dogecoin, Litecoin, Ripple, Monero, Ethereum, Cardano, Bitcoin Cash, Chainlink, and Binance due to their faster mean reversion and smaller volatility half-lives, and long-term investment in Bitcoin and Stellar due to their slower mean reversion and larger volatility half-lives. Recommendations include the use of heavy-tailed error distributions in modeling cryptocurrency returns and caution against excessive trading to mitigate market volatility
Keywords:
Cryptocurrency,
Money,
Mea Reversion,
Volatility Half-Life,
GARCH Model