Turn-of-the-month effect in cryptocurrencies

Abstract

Purpose This study examines the turn-of-the-month (TOM) effect in Bitcoin (BIT), Ethereum (ETH) and Litecoin (LIT) cryptocurrencies from August 2015 to August 2021.Design/methodology/approach Dummy regression model is used to examine the presence of the TOM effect and to test the efficiency of the cryptocurrency market. The characteristics of the returns during TOM days are compared with that of the non-non-TOM trading days. The authors also develop a trading strategy to earn abnormal returns using the TOM effect.Findings The authors show that TOM returns are positive and significantly higher than that of non-TOM returns. Interestingly, the authors empirically show that the TOM effect is not driven by the day-of-the-week (DOW) effect or the January effect. Based on the significant TOM effect, the authors formulate a trading strategy that annually outperforms the buy-and-hold strategy for BIT by 21.77% and for LIT by 47.10%. Finally, the results are robust to using a Generailzed Auto Regressive Conditional Heteroskedasticity (GARCH) (1,1) model and the January 2018 sell-off.Practical implications The results have important implications for both traders and investors. The findings suggest that the investors might be able to earn excess profits by timing their positions in BIT and LIT taking the advantage of the TOM effect.Originality/value First, the authors provide the only study to report the evidence of the TOM effect in three leading cryptocurrencies, viz., BIT, LIT and ETH. Second, the authors control for the DOW effect and the January effect while investigating the TOM effect in cryptocurrency market. Finally, this study develops a trading strategy based on which the investors can time the cryptocurrency markets as indicated by the pattern of the TOM effect during the studied time period.