Calendar Effects on Cryptocurrencies: Not so Straightforward

Abstract

This research investigates the existence of calendar anomalies on cryptocurrency markets with respect to calendar anomalies during 2010–2020. To account for potential clustering and non-normality in cryptocurrency returns, generalized autoregressive conditional heteroscedasticity (GARCH) regression with dummy variables is utilized. Two exploitable trading strategies are identified. First, it was found that Ethereum investors can generate abnormal returns in January. Second, abnormal profits can be generated from short-selling Litecoin on Mondays. Neither calendar anomaly is unique to the global COVID-19 pandemic. These results are robust after the considerations
of volatility clustering, non-normality, and changes in methodologies to detect the anomalies, and are consistent with the literature in stock markets.