Lennard Kraft

This study examines seasonal and momentum effects in Bitcoin markets, focusing on the period from April 2020 to December 2024. The analysis covers intraday, intraweek, and intrayear seasonal patterns, as well as short term momentum effects. The study also examined whether momentum continues into the next day after a “momentum day”. The end-of-the-month phenomenon is investigated as well.

Bitcoin exhibits significant intraday seasonality, indicating a relevant momentum effect. This effect occurs in the last four trading hours from 8 pm till the end of a day. Several intraday momentum strategies, buying this last four hours, were tested and delivered notable results. Conversely on a day after a “momentum day” results were mixed. In some market conditions, momentum persisted, while in others, a reversal was observed.

The intraweek analysis highlights a trend-behavior, but no period with significant outperformance. Mondays and Wednesdays have the strongest average returns. Intramonth analysis revealed a strong end-of-the-month effect, particularly between the 25th of one month and the 5th of the next. Intrayear findings indicate that Bitcoin reflects seasonal patterns observed in U.S. equities, such as the “Sell in May and Go Away” effect.

This study investigates Bitcoin’s price dynamics to explore its role as a financial asset and how it interacts with traditional markets. This investigation is crucial, as the connection between Bitcoin and the U.S. stock market has become increasingly relevant in recent years. The research demonstrates that since March 2020, Bitcoin has functioned as a risk-on asset, exhibiting optimal performance during favorable U.S. market conditions. I developed linear regression models, using the VIX Index and VX term structure, to demonstrate, how low and high VIX volatility regimes affect Bitcoin prices. For these two investigations, equity curves were plotted, pointing out, that investors could avoid Bitcoin’s sharp drawdowns, if they stayed out of the market during high VIX regimes. Furthermore, this research presents a valuation tool, based on the TQQQ ETF, to determine overbought and oversold areas. Bitcoin and the ETF demonstrate exhibit rangebound behavior, making this relationship to a useful timing component. Common statistical metrics are used to determine a normal distribution between TQQQ and Bitcoin, to prove the effectiveness of the tool.